Method and apparatus for generating a composite image using the difference of two images

Claim 2

An imaging system useful for compositing images, said system comprising: (a) a camera for capturing multiple images, including a first image of a subject plus background and a second image of the background only; (b) means for converting images into digital form; (c) means for subtracting the first and second images to generate a difference image; (d) a mask generating section for processing the difference image to generate a mask image, said mask generating section including: (i) a thresholding operation to reduce noise, wherein the thresholding operation includes a threshold that is set based on the noise of at least one of the captured images dependent upon the noise level of the camera that captured the images; (ii) means for processing the difference image including means for finding local zero-mean, low-variance regions; (iii) means responsive to the low-variance regions for estimating a noise variance; and (iv) means responsive to the noise variance for setting the threshold for the thresholding operation; and (e) means for applying the mask image to the first image to select an area in which the subject is located, whereby the selected subject area is used in the compositing of a new image.

Claim 3

The imaging system as claimed in claim 1 wherein the threshold set in the thresholding operation is based on the noise in the difference image.

Claim 4

The imaging system as claimed in claim 1 wherein the thresholding operation is stored in a lookup table.

Claim 5

The imaging system as claimed in claim 1 wherein an output of the thresholding operation is applied to a despeckling filter to eliminate random noise in the mask image.

Claim 6

The imaging system as claimed in claim 1 wherein a user interface allows a user of the system to manually clean the mask image.

Claim 7

The imaging system as claimed in claim 5 wherein the user interface provides a user controlled tool to manually clean the mask image.

Claim 8

The imaging system as claimed in claim 1 wherein said camera captures a plurality of second images of the background and said system further includes means for averaging the plurality of second images in order to minimize noise in the background.

Claim 9

The imaging system as claimed in claim 1 further comprising: means responsive to said applying means for extracting a subject image from the first image; means for selecting a different background image; and means for generating a composite image from the subject image and the different background image.

Claim 10

The imaging system as claimed in claim 1 further comprising: means responsive to said applying means for extracting a subject image from the first image; means for processing the subject image; and means for generating a composite image from the processed subject image and the second image of the background only.

Claim 11

The imaging system as claimed in claim 1 further comprising: means responsive to said applying means for extracting a subject image from the first image; means for processing the second image of the background only; means for generating a composite image from the subject image and the processed second image.

Claim 12

An electronic imaging system useful for compositing images, said system comprising: (a) an electronic camera for capturing multiple color images each composed of separate color records, including a first color image of a subject plus background and a second color image of the background only; (b) means for subtracting the separate color records composing said first and second images to generate a difference image for each color record; (c) a processing section for processing the separate difference images to generate a mask image for each color record, said processing section including: (i) a noise reduction algorithm to reduce noise in each mask image, wherein the noise reduction algorithm includes a separate threshold that is applied to image values in each of the difference images corresponding to each of the color records, and each threshold is set to be proportional to at least one of the captured images dependent upon the noise level of the color channels of the camera that captured the color images; (ii) means for processing each of the difference images including means for finding local zero-mean, low-variance regions in the separate difference images; (iii) means responsive to the low-variance regions for estimating a noise variance for each of the color records; and (iv) means responsive to the noise variance for setting the separate threshold for each of the thresholding operations for each of the color records; and (d) means for applying the separate mask images to the first image to select an area in which the subject is located, whereby the selected subject area is used in the compositing of a new image.

Claim 13

An electronic imaging system useful for compositing images, said system comprising: (a) an electronic camera for capturing multiple color images each composed of separate color records, including a first color image of a subject plus background and a second color image of the background only; (b) means for subtracting the separate color records composing said first and second images to generate a difference image for each color record; (c) a processing section for processing the separate difference images to generate a mask image for each color record, said processing section including: (i) a noise reduction algorithm to reduce noise in each mask image, wherein the noise reduction algorithm includes a separate threshold that is applied to image values in each of the difference images corresponding to each of the color records, and each threshold is set to be proportional to at least one of the captured images dependent upon the noise level of the color channels of the camera that captured the color images; (ii) means for processing each of the difference images including means for finding local zero-mean, low-variance regions in the separate difference images; (iii) means responsive to the low-variance regions for estimating a noise variance for each of the color records; and (iv) means responsive to the noise variance for setting the separate threshold for each of the thresholding operations for each of the color records; and (d) means for applying the separate mask images to the first image to select an area in which the subject is located, whereby the selected subject area is used in the compositing of a new image.

Claim 14

The electronic imaging system as claimed in claim 11 wherein each thresholding operation includes a threshold that is set to be a function of proportional to the noise in the corresponding difference image.

Claim 15

The electronic imaging system as claimed in claim 11 wherein the thresholding operations are stored in one or more lookup tables.

Claim 16

The electronic imaging system as claimed in claim 13 wherein an output of each thresholding operation is applied to a despeckling filter to eliminate random noise in the difference image in each color record.

Claim 17

The electronic imaging system as claimed in claim 14 wherein an output of each despeckling filter is combined in an OR operation to generate a binary mask.

Claim 18

The electronic imaging system as claimed in claim 15 wherein a user interface allows a user of the system to operate upon the binary mask and manually clean the binary mask.

Claim 19

The electronic imaging system as claimed in claim 16 wherein the user interface provides a user controlled tool to manually clean the binary mask.

Claim 20

The electronic imaging system as claimed in claim 11 wherein said camera captures a plurality of second images of the background and said system further includes means for averaging the plurality of second images in order to minimize noise in the background.

Claim 21

The electronic imaging system as claimed in claim 11 further comprising: means responsive to said applying means for extracting a subject image from the first image; means for selecting a different background image; and means for generating a composite image from the subject image and the different background image.

Claim 22

An electronic imaging system useful for compositing images, said system comprising: (a) an electronic camera for capturing a motion sequence of first images of a subject plus background and a second image of the background only; (b) means for subtracting each of the first images and the second image to generate a sequence of difference images; (c) a noise processing and mask generating section including: (i) a noise reduction thresholding operation to reduce noise in the difference images in order to generate a sequence of foreground mask images, wherein the thresholding operation includes a threshold that is set to be a function of the noise in the electronic camera; (ii) means for processing at least one difference image including means for finding local zero-mean, low-variance regions in the at least one difference image; (iii) means responsive to the low-variance regions for estimating a noise variance; and (iv) means responsive to the noise variance for setting the threshold for the thresholding operation; and (d) means for applying the mask images to the sequence of first images to select an area in which the subject is located in each of the first images, whereby the selected subject areas are used in the compositing of a new sequence of images.

Claim 23

An electronic imaging system useful for compositing images, said system comprising: (a) an electronic camera for capturing a motion sequence of first images of a subject plus background and a second image of the background only; (b) means for subtracting each of the first images and the second image to generate a sequence of difference images; (c) a noise processing and mask generating section including: (i) a noise reduction thresholding operation to reduce noise in the difference images in order to generate a sequence of foreground mask images, wherein the thresholding operation includes a threshold that is set to be a function of the noise in the electronic camera; (ii) means for processing at least one difference image including means for finding local zero-mean, low-variance regions in the at least one difference image; (iii) means responsive to the low-variance regions for estimating a noise variance; and (iv) means responsive to the noise variance for setting the threshold for the thresholding operation; and (d) means for applying the mask images to the sequence of first images to select an area in which the subject is located in each of the first images, whereby the selected subject areas are used in the compositing of a new sequence of images.

Claim 24

The electronic imaging system as claimed in claim 20 wherein the thresholding operation includes a threshold that is set to be a function of the noise in the difference images.

Claim 25

The electronic imaging system as claimed in claim 20 wherein the thresholding operation is stored in a lookup table.

Claim 26

The electronic imaging system as claimed in claim 20 wherein an output of the thresholding operation is applied to a despeckling filter to eliminate random noise in the foreground mask images.

Claim 27

The electronic imaging system as claimed in claim 20 wherein a user interface allows a user of the system to manually clean the foreground mask images.

Claim 28

The electronic imaging system as claimed in claim 24 wherein the user interface provides a user controlled tool to manually clean the foreground mask images.

Claim 29

The electronic imaging system as claimed in claim 20 wherein said camera captures a plurality of second images of the background and said system further includes means for averaging the plurality of second images in order to minimize noise in the background.

Claim 30

The electronic imaging system as claimed in claim 20 further comprising: means responsive to said applying means for extracting a subject image from each of the first images; means for selecting a different background image; and means for generating a composite image for each of the sequence of first images from the corresponding subject image and the different background image.

Claim 31

The electronic imaging system as claimed in claim 20 wherein said noise processing and mask generating section uses a motion tracking algorithm to calculate some of the foreground mask images from the difference images.

Claim 32

The electronic imaging system as claimed in claim 20 wherein said electronic camera captures a sequence of color images each composed of separate color records, including a sequence of first color images of a subject plus background and a second color image of the background only, and said subtracting means subtracts the separate color records composing said first and second images to generate a difference image for each color record.

Claim 33

An electronic imaging method useful for compositing images, said method comprising the steps of: (a) capturing multiple images, including a first image of a subject plus background and a second image of the background only; (b) subtracting the first and second images to generate a difference image; (c) setting a noise reduction threshold as a function of noise in the difference image, including the steps of: (i) processing the difference image to find local zero-mean, low-variance regions; (ii) estimating, responsive to the low-variance regions, a noise variance; and (iii) setting, responsive to the noise variance, the noise reduction threshold; (d) applying the noise reduction threshold to the difference image to generate a mask image; and (e) applying the mask image to the first image to select an area in which the subject is located, whereby the selected subject area is used in the compositing of a new image.

Claim 34

An electronic imaging method useful for compositing images, said method comprising the steps of: (a) capturing multiple images, including a first image of a subject plus background and a second image of the background only; (b) subtracting the first and second images to generate a difference image; (c) setting a noise reduction threshold as a function of noise in the difference image, including the steps of: (i) processing the difference image to find local zero-mean, low-variance regions; (ii) estimating, responsive to the low-variance regions, a noise variance; and (iii) setting, responsive to the noise variance, the noise reduction threshold; (d) applying the noise reduction threshold to the difference image to generate a mask image; and (e) applying the mask image to the first image to select an area in which the subject is located, whereby the selected subject area is used in the compositing of a new image.

Claim 35

The electronic imaging method as claimed in claim 30 wherein the noise reduction threshold operation is stored in a lookup table.

Claim 36

The electronic imaging method as claimed in claim 30 wherein an output of the noise reduction threshold operation is applied to a despeckling filter to eliminate random noise in the mask image.

Claim 37

The electronic imaging method as claimed in claim 32 further including the step of operating upon the output of the noise reduction threshold operation and cleaning the mask image.

Claim 38

The electronic imaging method as claimed in claim 30 further including the steps of capturing a plurality of second images of the background and averaging the plurality of second images in order to minimize noise in the background.

Claim 39

The electronic imaging method as claimed in claim 30 further comprising the steps of: extracting a subject image from the first image; selecting a different background image; and generating a composite image from the subject image and the different background image.