Device and method for smoothing picture signal, device and method for encoding picture and device and method for decoding picture

Claim 2

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted mean of a target pixel and its surrounding pixels, including: means for detecting a difference between the target pixel and the surrounding pixels; means for taking an absolute value that represents a magnitude of the difference; means for adding a non-negative value to the absolute value that represents a magnitude of the difference thus detected, wherein said non-negative value is equal for all surrounding pixels; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixels value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; and means for smoothing the target pixel in accordance with the normalized sum (A).

Claim 3

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; and means for appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to roughness of quantizing parameter of each block, which is represented by quantizing characteristics utilized in encoding or decoding each block of a picture signal.

Claim 4

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; and means for appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to roughness of quantizing parameter of each block, which is represented by quantizing characteristics utilized in encoding or decoding each block of a picture signal.

Claim 5

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for computing block activity indicating a change degree of a pixel in each block; and means for appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to the block activity.

Claim 6

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for computing block activity indicating a change degree of a pixel in each block; and means for appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to the block activity.

Claim 7

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for detecting whether a target pixel exists on a block boundary; means that if the target pixel exits on a block boundary, detects a block activity of a block adjacent to the block boundary, i.e., adjacent block activity; and means for appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to the adjacent block activity.

Claim 8

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for detecting whether a target pixel exists on a block boundary; means that if the target pixel exits on a block boundary, detects a block activity of a block adjacent to the block boundary, i.e., adjacent block activity; and means for appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to the adjacent block activity.

Claim 9

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for mixing a signal value after the smoothing process by block and a signal value prior to the smoothing process; and means for appropriately changing the mixing ratio according to a roughness of quantizing characteristics utilized in encoding and decoding each block of a picture signal.

Claim 10

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for mixing a signal value after the smoothing process by block and a signal value prior to the smoothing process; and means for appropriately changing the mixing ratio according to a roughness of quantizing characteristics utilized in encoding and decoding each block of a picture signal.

Claim 11

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for mixing a signal value after the smoothing process by block and a signal value prior to the smoothing process; and means for appropriately changing the mixing ratio according to each block activity in a picture signal.

Claim 12

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for mixing a signal value after the smoothing process by block and a signal value prior to the smoothing process; and means for appropriately changing the mixing ratio according to each block activity in a picture signal.

Claim 13

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for mixing a signal value after smoothing process by block and signal value prior to the smoothing process; means for detecting whether a target pixel exists on a block boundary; and means that if the target pixel exists on a block boundary, appropriately changes the mixing ratio according to an adjacent block activity of a block adjacent to the target pixel.

Claim 14

A picture signal smoothing device that performs a smoothing process by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factors; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for mixing a signal value after smoothing process by block and signal value prior to the smoothing process; means for detecting whether a target pixel exists on a block boundary; and means that if the target pixel exists on a block boundary, appropriately changes the mixing ratio according to an adjacent block activity of a block adjacent to the target pixel.

Claim 15

A picture signal smoothing device that performs a filtering of a sampled and digitized picture signal by a smoothing filter, comprising: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factor values; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for detecting whether a target pixel exists on a block boundary; means for performing a filtering whose edge preserving property is weakened, if the target pixel exists on a block boundary; means for performing a filtering having edge preserving property if the target pixel exists inside a block; means for computing an activity of the target pixel from the target pixel value and the surrounding pixel value; and means for increasing the strength of the edge preserving property of a filter responsive to the activities of the target pixel if the target pixel exists on a block boundary and decreasing the strength of the edge preserving property of a filter responsive to the activities of the target pixel if the target pixel exists inside the block boundary.

Claim 16

A picture signal smoothing device that performs a filtering of a sampled and digitized picture signal by a smoothing filter, comprising: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value to the absolute value of the difference thus detected; means for taking an inverse number of the value after the adding and thereby generating a weighting factor; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factors alone; means for normalizing the sum (A) by dividing the same by the sum (B) of the weighting factor values; means for smoothing the target pixel in accordance with the normalized sum (A); means for dividing a picture to be filtered into blocks; means for detecting whether a target pixel exists on a block boundary; means for performing a filtering whose edge preserving property is weakened, if the target pixel exists on a block boundary; means for performing a filtering having edge preserving property if the target pixel exists inside a block; means for computing an activity of the target pixel from the target pixel value and the surrounding pixel value; and means for increasing the strength of the edge preserving property of a filter responsive to the activities of the target pixel if the target pixel exists on a block boundary and decreasing the strength of the edge preserving property of a filter responsive to the activities of the target pixel if the target pixel exists inside the block boundary.

Claim 17

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by a smoothing filter, comprising: detecting a difference between a target pixel and its surrounding pixels; taking an absolute value that represents a magnitude of the difference; adding a non-negative value (m) to the absolute value that represents a magnitude of the difference thus detected, wherein said non-negative value is equal for all surrounding pixels; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixels value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; and smoothing the target pixel in accordance with the normalized sum (A).

Claim 18

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by a smoothing filter, comprising: detecting a difference between a target pixel and its surrounding pixels; taking an absolute value that represents a magnitude of the difference; adding a non-negative value (m) to the absolute value that represents a magnitude of the difference thus detected, wherein said non-negative value is equal for all surrounding pixels; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixels value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; and smoothing the target pixel in accordance with the normalized sum (A).

Claim 19

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal byweighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; and appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to the roughness of quantizing parameter of each block, which is represented by quantizing characteristics utilized in encoding and decoding each block of a picture signal.

Claim 20

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal byweighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; and appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to the roughness of quantizing parameter of each block, which is represented by quantizing characteristics utilized in encoding and decoding each block of a picture signal.

Claim 21

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; computing a block activity indicating a change degree of a pixel in each block; and appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to the block activity.

Claim 22

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; computing a block activity indicating a change degree of a pixel in each block; and appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to the block activity.

Claim 23

A picture signal smoothing method in which a smoothing process is performed byapplying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; detecting whether a target pixel exists on a block boundary; if the target pixel exists on a block boundary, appropriately changing a block activity of a block adjacent to the block boundary, i.e., adjacent block activity; and appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to the adjacent block activity.

Claim 24

A picture signal smoothing method in which a smoothing process is performed byapplying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; detecting whether a target pixel exists on a block boundary; if the target pixel exists on a block boundary, appropriately changing a block activity of a block adjacent to the block boundary, i.e., adjacent block activity; and appropriately changing a non-negative value to be added to the difference value between the target pixel and the surrounding pixel, according to the adjacent block activity.

Claim 25

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; mixing a signal value after the smoothing process by block and a signal value prior to the smoothing process; and appropriately changing the mixing ratio according to a roughness of quantizing parameter of each block, which is represented by quantizing characteristics utilized in encoding and decoding each block of a picture signal.

Claim 26

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; mixing a signal value after the smoothing process by block and a signal value prior to the smoothing process; and appropriately changing the mixing ratio according to a roughness of quantizing parameter of each block, which is represented by quantizing characteristics utilized in encoding and decoding each block of a picture signal.

Claim 27

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; mixing a signal value after the smoothing process by block and a signal value prior to the smoothing process; and appropriately changing the mixing ratio according to each block activity in picture signal.

Claim 28

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; mixing a signal value after the smoothing process by block and a signal value prior to the smoothing process; and appropriately changing the mixing ratio according to each block activity in picture signal.

Claim 29

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; mixing a signal value after the smoothing process by block and a signal value prior to the smoothing process; detecting whether a target pixel exists on a block boundary; and if the target pixel exists on a block boundary, appropriately changing the mixing ratio according to an adjacent block activity of a block adjacent to the target pixel.

Claim 30

A picture signal smoothing method in which a smoothing process is performed by applying a smoothing filter that performs a filtering of a sampled and digitized picture signal by weighted means of a target pixel and its surrounding pixel, including the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of values weighting factors alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; mixing a signal value after the smoothing process by block and a signal value prior to the smoothing process; detecting whether a target pixel exists on a block boundary; and if the target pixel exists on a block boundary, appropriately changing the mixing ratio according to an adjacent block activity of a block adjacent to the target pixel.

Claim 31

A picture signal smoothing method in which a filtering of a sampled and digitized picture signal is performed by a smoothing filter, comprising the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; normalizing the sum (A) by dividing the same by the sum (B) of the weighting factor values; and smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; detecting whether a target pixel exists on a block boundary; performing a filtering whose edge preserving property is weakened, if the target pixel exists on a block boundary; performing a filtering having edge preserving property if the target pixel exists inside a block; computing an activity of the target pixel from a target pixel value and its surrounding pixel value; and increasing the strength of edge preserving property of a filter according to the activity of the target pixel if the target pixel exists on a block boundary and decreasing the strength of edge preserving of a filter according to the activity of the target pixel if the target pixel exists inside the block boundary and decreasing the strength of edge preserving of a filter according to the activity of the target pixel exists inside the block boundary.

Claim 32

A picture signal smoothing method in which a filtering of a sampled and digitized picture signal is performed by a smoothing filter, comprising the steps of: detecting a difference between a target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; normalizing the sum (A) by dividing the same by the sum (B) of the weighting factor values; and smoothing the target pixel in accordance with the normalized sum (A); dividing a picture to be filtered into blocks; detecting whether a target pixel exists on a block boundary; performing a filtering whose edge preserving property is weakened, if the target pixel exists on a block boundary; performing a filtering having edge preserving property if the target pixel exists inside a block; computing an activity of the target pixel from a target pixel value and its surrounding pixel value; and increasing the strength of edge preserving property of a filter according to the activity of the target pixel if the target pixel exists on a block boundary and decreasing the strength of edge preserving of a filter according to the activity of the target pixel if the target pixel exists inside the block boundary and decreasing the strength of edge preserving of a filter according to the activity of the target pixel exists inside the block boundary.

Claim 33

A picture encoding device in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame encoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: means for detecting a difference between the target pixel and the surrounding pixels; means for taking an absolute value of the difference; means for adding a non-negative value (m) to the absolute value of the difference thus detected; means for taking an inverse number (weighting factor) of the value after the adding; means for multiplying a value after taking of the inverse number, and the surrounding pixels value together; means for taking a sum (A) of the value after the multiplying; means for taking a sum (B) of the inverse number (the weighting factor) alone; means for dividing the sum (A) by the sum (B) of the weighting factor and then normalizing (A/B); a filter having edge preserving property is employed as the above filter; and a picture signal after being processed in a picture signal smoothing device is input to the frame memory storing a picture signal of a preceding frame, as a local decoding signal.

Claim 34

A picture encoding device in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame encoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: means for detecting a difference between the target pixel and the surrounding pixels; means for taking an absolute value of the difference; means for adding a non-negative value (m) to the absolute value of the difference thus detected; means for taking an inverse number (weighting factor) of the value after the adding; means for multiplying a value after taking of the inverse number, and the surrounding pixels value together; means for taking a sum (A) of the value after the multiplying; means for taking a sum (B) of the inverse number (the weighting factor) alone; means for dividing the sum (A) by the sum (B) of the weighting factor and then normalizing (A/B); a filter having edge preserving property is employed as the above filter; and a picture signal after being processed in a picture signal smoothing device is input to the frame memory storing a picture signal of a preceding frame, as a local decoding signal.

Claim 35

A picture encoding device according to claim 17 including: means for dividing a picture into blocks and computing an inter-frame difference by block; means for determining whether the difference exceeds a certain threshold value; and means that only when the difference exceeds the threshold value, a corresponding block is processed by a filter having edge preserving property.

Claim 36

A picture encoding device in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame encoding is performed by a corresponding pixel between an input signal and the frame memory comprising: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value (m) to the absolute value of the difference thus detected; means for taking an inverse number (weighting factor) of the value after the adding; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of the value after the multiplying; means for taking a sum (B) of a plurality of weighting factor values alone; and means for dividing the sum (A) by the sum (B) of the weighting factor and then normalizing (A/B); smoothing the target pixel in accordance with the normalized sum (A); means for computing a motion vector; means for reading a picture value stored in the frame memory with consideration of the motion vector; and a filter having edge preserving property, which filter processes an output signal of the frame memory.

Claim 37

A picture encoding device in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame encoding is performed by a corresponding pixel between an input signal and the frame memory comprising: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value (m) to the absolute value of the difference thus detected; means for taking an inverse number (weighting factor) of the value after the adding; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of the value after the multiplying; means for taking a sum (B) of a plurality of weighting factor values alone; and means for dividing the sum (A) by the sum (B) of the weighting factor and then normalizing (A/B); smoothing the target pixel in accordance with the normalized sum (A); means for computing a motion vector; means for reading a picture value stored in the frame memory with consideration of the motion vector; and a filter having edge preserving property, which filter processes an output signal of the frame memory.

Claim 38

A picture decoding device in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame decoding is performed by a corresponding pixel between an input signal and the frame memory comprising: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value (m) to the absolute value of the difference thus detected; means for taking an inverse number (weighting factor) of the value after the adding; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factor values alone; and means for dividing the sum (A) by the sum (B) of the weighting factor values and then normalizing (A/B); a filter having edge preserving property is employed as the above filter; and a picture signal after being processed in a picture signal smoothing device is input to a frame memory storing a picture signal of the previous frame, as a decoding signal.

Claim 39

A picture decoding device in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame decoding is performed by a corresponding pixel between an input signal and the frame memory comprising: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value (m) to the absolute value of the difference thus detected; means for taking an inverse number (weighting factor) of the value after the adding; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factor values alone; and means for dividing the sum (A) by the sum (B) of the weighting factor values and then normalizing (A/B); a filter having edge preserving property is employed as the above filter; and a picture signal after being processed in a picture signal smoothing device is input to a frame memory storing a picture signal of the previous frame, as a decoding signal.

Claim 40

A picture decoding device according to claim 20 including: means for dividing a picture into blocks and computing an inter-frame difference by block; means for determining whether the difference exceeds a certain threshold value; and means that only when the difference exceeds the threshold value, a corresponding block is processed by a filter having edge preserving property.

Claim 41

A picture decoding device in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one filed, and an inter-frame decoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value (m) to the absolute value of the difference thus detected; means for taking an inverse number (weighting factor) of the value after the adding; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factor values alone; and means for dividing the sum (A) by the sum (B) of the weighting factor values and then normalizing (A/B); means for computing a motion vector; means for reading a picture value stored in the frame memory with consideration of the motion vector; and a filter having edge preserving property, which filter processes an output signal of the frame memory.

Claim 42

A picture decoding device in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one filed, and an inter-frame decoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: means for detecting a difference between the target pixel and the surrounding pixel; means for taking an absolute value of the difference; means for adding a non-negative value (m) to the absolute value of the difference thus detected; means for taking an inverse number (weighting factor) of the value after the adding; means for multiplying a value after taking of the inverse number, and the surrounding pixel value together; means for taking a sum (A) of a plurality of values after the multiplying; means for taking a sum (B) of a plurality of weighting factor values alone; and means for dividing the sum (A) by the sum (B) of the weighting factor values and then normalizing (A/B); means for computing a motion vector; means for reading a picture value stored in the frame memory with consideration of the motion vector; and a filter having edge preserving property, which filter processes an output signal of the frame memory.

Claim 43

A picture encoding method in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame encoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: detecting a difference between the taraet pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number (weighting factor) of a value after the adding; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; dividing the sum (A) by the sum (B) of the weighting factor values and then normalizing (A/B); performing a process using a filter having edge preserving property; and inputting a picture signal after being processed in a picture signal smoothing device to the frame memory storing a picture signal of a preceding frame, as a local decoding signal.

Claim 44

A picture encoding method in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame encoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: detecting a difference between the taraet pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number (weighting factor) of a value after the adding; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; dividing the sum (A) by the sum (B) of the weighting factor values and then normalizing (A/B); performing a process using a filter having edge preserving property; and inputting a picture signal after being processed in a picture signal smoothing device to the frame memory storing a picture signal of a preceding frame, as a local decoding signal.

Claim 45

A picture encoding method according to claim 23 including: dividing a picture into blocks and computing an inter-frame difference by block; determining whether the difference exceeds a certain threshold value; and only when the difference exceeds the threshold value, a corresponding block is processed by a filter having edge preserving property.

Claim 46

A picture encoding method in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame encoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: detecting a difference between the target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number (weighting factor) of a value after the adding; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; dividing the sum (A) by the sum (B) of the weighting factor values and then normalizing (A/B); computing a motion vector; reading a picture value stored in the frame memory with consideration of the motion vector; and processing an output signal of the frame memory by using a filter having edge preserving property.

Claim 47

A picture encoding method in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame encoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: detecting a difference between the target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number (weighting factor) of a value after the adding; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; dividing the sum (A) by the sum (B) of the weighting factor values and then normalizing (A/B); computing a motion vector; reading a picture value stored in the frame memory with consideration of the motion vector; and processing an output signal of the frame memory by using a filter having edge preserving property.

Claim 48

A picture encoding method in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame encoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: detecting a difference between the target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number (weighting factor) of a value after the adding; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; dividing the sum (A) by the sum (B) of the weighting factor values and then normalizing (A/B); performing a process using a filter having edge preserving property; and inputting a picture signal after being processed in a picture signal smoothing device to the frame memory storing a picture signal of a preceding frame, as a decoding signal.

Claim 49

A picture encoding method in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one field, and an inter-frame encoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: detecting a difference between the target pixel and its surrounding pixel; taking an absolute value of the difference; adding a non-negative value (m) to the absolute value of the difference thus detected; taking an inverse number (weighting factor) of a value after the adding; multiplying a value after taking of the inverse number and the surrounding pixel value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; dividing the sum (A) by the sum (B) of the weighting factor values and then normalizing (A/B); performing a process using a filter having edge preserving property; and inputting a picture signal after being processed in a picture signal smoothing device to the frame memory storing a picture signal of a preceding frame, as a decoding signal.

Claim 50

A picture decoding method according to claim 26 further comprising: dividing a picture into blocks and computing an inter-frame difference by block; determining whether the difference exceeds a certain threshold value; and only when the difference exceeds a certain threshold value, a corresponding block is processed by a filter having an edge preserving property.

Claim 51

A picture decoding method in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one filed, and an inter-frame decoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: detecting a difference between a target pixel and its surrounding pixels; taking an absolute value that represents a magnitude of the difference; adding a non-negative value (m) to the absolute value that represents a magnitude of the difference thus detected, wherein said non-negative value is equal for all surrounding pixels; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixels value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; and smoothing the target pixel in accordance with the normalized sum (A); computing a motion vector; reading a picture value stored in te frame memory with consideration of a motion vector; and processing an output signal of the frame memory by using a filter having edge preserving property.

Claim 52

A picture decoding method in which a filter is attached to a decoding loop having a frame memory for storing a signal that corresponds to either of at least one frame and at least one filed, and an inter-frame decoding is performed by a corresponding pixel between an input signal and the frame memory, comprising: detecting a difference between a target pixel and its surrounding pixels; taking an absolute value that represents a magnitude of the difference; adding a non-negative value (m) to the absolute value that represents a magnitude of the difference thus detected, wherein said non-negative value is equal for all surrounding pixels; taking an inverse number of a value after the adding thereby generating a weighting factor; multiplying a value after taking of the inverse number and the surrounding pixels value together; taking a sum (A) of a plurality of values after the multiplying; taking a sum (B) of a plurality of weighting factor values alone; normalizing the sum (A) by dividing same by the sum (B) of the weighting factors; and smoothing the target pixel in accordance with the normalized sum (A); computing a motion vector; reading a picture value stored in te frame memory with consideration of a motion vector; and processing an output signal of the frame memory by using a filter having edge preserving property.