Method and apparatus for detecting and processing noisy edges in image detail enhancement

Claim 3

The method of claim 1 wherein in step (a) detecting image pixels that belong to an image edge, further includes the steps of detecting image pixels that belong to a horizontal or vertical image edge.

Claim 4

The method of claim 1 wherein in step (a) detecting image pixels that belong to an image edge, further includes the steps of detecting image pixels that belong to a noisy and sharp horizontal or vertical image edge.

Claim 5

The method of claim 1 , wherein in step (b) generating an unsharp image signal further includes the steps of: determining mean values of pixels in a two-dimensional window including the selected pixel; and using said mean values in a filtering process to obtain an unsharp image signal at the selected pixel location.

Claim 6

The method of claim 1 , where in step (a) detecting image pixels that belong to an image edge further includes the steps of: selecting a pixel; and based on the selected pixel and its neighboring pixels in a two-dimensional window, determining if the pixel belongs to an image edge; wherein each pixel is checked separately to determine if it belongs to an image edge.

Claim 7

The method of claim 1 , further comprising the steps of: if a pixel is determined to belong to an image edge, then determining if the edge is noisy.

Claim 8

The method of claim 5 , wherein detecting image pixels that belongs to an essentially vertical image edge further includes the steps of: selecting at least W columns of pixels centered with the selected pixel, wherein each columns includes H pixels; determining the mean value m of the pixels in each column, thereby generating W mean values; determining the variance value σ of the pixels in each column, thereby generating W variance values; and based on the W mean values and the W variance values, determining if the selected pixel belongs to a vertical image edge.

Claim 9

The method of claim 7 , wherein: the mean values m j are determined according to the relation: m j = 1 H ⁢ ∑ i = - H - 1 2 H - 1 2 ⁢ I i , j , j = - 1 , 0 , 1 , the variance values σ j are determined according to the relation: σ j = 1 H ⁢ ∑ i = - H - 1 2 H - 1 2 ⁢  I i , j ⁢ ⁢ m j  ⁢ , ⁢ j = - 1 , 0 , 1 , where I i,j is the value of a pixel located at row i and column j in the window, such that the row and column index of the selected pixel is 0.

Claim 10

The method of claim 7 , wherein the steps of detecting if the selected pixel belongs to an image edge, further includes the steps of: determining if |m 0 −m j |>max(σ 0 ,σ j ), j=−1 or 1; determining if max(|m 0 −m 1 |,|m 0 −m −1 |)≧T m ; where T m is a predetermined threshold value; such that if |m 0 −m j |>max(σ 0 ,σ j ), j=−1 or 1, and if max(|m 0 −m 1 |,|m 0 −m −1 |)≧T m , then the selected pixel is considered as a pixel in an image edge.

Claim 11

The method of claim 7 , further comprising the steps of: if the selected pixel is determined to belong to an image edge, then determining if the edge is noisy.

Claim 12

The method of claim 10 , wherein the step of checking if the edge is noisy further includes the steps of: generating binary pattern data b from pixels in column 0 of the window according to the relation: b i , 0 = { 0 if ⁢ ⁢ I i , 0 < m 0 1 if ⁢ ⁢ I i , 0 ≥ m 0 ⁢ ⁢ i = - H - 1 2 , … ⁢ , 0 , … ⁢ , H - 1 2 ; based on the binary pattern data, generating a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ i = - H - 1 2 H - 1 2 - 1 ⁢  b i , 0 - b i + 1 , 0  ; comparing the count N to a predetermined threshold value T N , 0<T N <H−1; if the count N is not less than T N , then the edge is considered as noisy.

Claim 13

The method of claim 10 , wherein the step of checking if the edge is noisy further includes the steps of: generating binary pattern data b from pixels in column 0 of the window according to the relation: b i , 0 = { 0 if ⁢ ⁢ I i , 0 < m 0 1 if ⁢ ⁢ I i , 0 ≥ m 0 ⁢ ⁢ i = - H - 1 2 , … ⁢ , 0 , … ⁢ , H - 1 2 ; based on the binary pattern data, generating a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ i = - H - 1 2 H - 1 2 - 1 ⁢ ( b i , 0 ⊕ ⁢ b i + 1 , 0 ) ; comparing the count N to a predetermined threshold value T N , 0<T N <H−1; if the count N is not less than T N , then the edge is considered as noisy.

Claim 14

The method of claim 4 , wherein: said window comprises a W ƒ by H ƒ rectangular window centered with the selected pixel; if the image edge direction is vertical: determining the mean values further includes the step of determining the mean values of the pixels in each column of the window, wherein there are a total of W ƒ such mean values; performing the filtering process further includes the steps of using a low pass filter (LPF) on the mean values to obtain an unsharp image signal at the selected pixel location.

Claim 15

The method of claim 13 , wherein performing the filtering process further includes the steps of: selecting a one dimensional (1D) LPF, such that W ƒ is equal to the length of the LPF; applying the LPF to the mean values to obtain the unsharp signal.

Claim 16

The method of claim 13 , wherein performing the filtering process further includes the steps of: selecting a two dimensional (2D) LPF, such that W ƒ and H ƒ are equal to the horizontal and vertical length of the LPF, respectively; generating a two-dimensional data array of the size W ƒ by H ƒ , wherein the data in each column of the array are all set to the corresponding pixel mean value of the same column in the W ƒ by H 71 window of neighboring pixels; and applying the LPF to the array data to obtain the unsharp signal.

Claim 17

The method of claim 5 , wherein detecting image pixels that belong to an essentially horizontal image edge further includes the steps of: selecting at least H rows of pixels centered with the selected pixel, wherein each row includes W pixels; determining the mean value of the pixels in each row, thereby generating H mean values; determining the variance value of the pixels in each row, thereby generating H variance values; based on the H mean values and the H variance values, determining if the selected pixel belongs to a horizontal image edge.

Claim 18

The method of claim 16 , wherein: the mean values m i are determined according to the relation: m i = 1 W ⁢ ∑ j = - W - 1 2 W - 1 2 ⁢ I i , j ⁢ ⁢ i = - 1 , 0 , 1 , the variance values σ i are determined according to the relation: σ i = 1 W ⁢ ∑ j = - W - 1 2 W - 1 2 ⁢  I i , j ⁢ ⁢ m i  ⁢ ⁢ i = - 1 , 0 , 1 , where I i,j is the value of a pixel located at row i and column j in the window, such that the row and column index of the selected pixel is 0.

Claim 19

The method of claim 16 , wherein the steps of detecting if the selected pixel belongs to an image edge, further includes the steps of: determining if |m 0 −m i |>max(σ 0 ,σ i ), i=−1 or 1; determining if max(|m 0 −m i |,|m 0 −m −1 |)≧T m ; where T m is a predetermined threshold value; such that if |m 0 −m i |>max(τ 0 ,σ i ), i=−1 or 1, and if max(|m 0 −m 1 |,|m 0 −m −1 |)≧T m , then the selected pixel is considered as a pixel in an image edge.

Claim 20

The method of claim 16 , further comprising the steps of: if the selected pixel is determined to belong to an image edge, then determining if the edge is noisy.

Claim 21

The method of claim 19 , wherein the step of checking if the edge is noisy further includes the steps of: generating binary pattern data b from pixels in row 0 of the window according to the relation: b 0 , j = { 0 if ⁢ ⁢ I 0 , j < m 0 1 if ⁢ ⁢ I 0 , j ≥ m 0 ⁢ ⁢ j = - W - 1 2 , … ⁢ , 0 , … ⁢ , W - 1 2 ; based on the binary pattern data, generating a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ j = - W - 1 2 W - 1 2 - 1 ⁢  b 0 , j - b 0 , j + 1  ; comparing the count N to a predetermined threshold value T N , 0<T N <W−1; if the count N is not less than T N , then the edge is considered as noisy.

Claim 22

The method of claim 19 , wherein the step of checking if the edge is noisy further includes the steps of: generating binary pattern data b from pixels in row 0 of the window according to the relation: b 0 , j = { 0 if ⁢ ⁢ I 0 , j < m 0 1 if ⁢ ⁢ I 0 , j ≥ m 0 ⁢ ⁢ j = - W - 1 2 , … ⁢ , 0 , … ⁢ , W - 1 2 ; based on the binary pattern data, generating a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ j = - W - 1 2 W - 1 2 - 1 ⁢ ( b 0 , j ⊕ b 0 , j + 1 ) ; comparing the count N to a predetermined threshold value T N , 0<T N <W−1; if the count N is not less than T N , then the edge is considered as noisy.

Claim 23

The method of claim 4 , wherein: said window comprises a W ƒ by H ƒ rectangular window centered with the selected pixel; if the image edge direction is horizontal: determining the mean values further includes the step of determining the mean values of the pixels in each row of the window, wherein there are a total of H ƒ such mean values; performing the filtering process further includes the steps of using a low pass filter (LPF) on the mean values to obtain an unsharp image signal at the selected pixel location.

Claim 24

The method of claim 22 , wherein performing the filtering process further includes the steps of: selecting a one dimensional (1D) LPF, such that H ƒ is equal to the length of the LPF; applying the LPF to the mean values to obtain the unsharp signal.

Claim 25

The method of claim 22 , wherein performing the filtering process further includes the steps of: selecting a two dimensional (2D) LPF, such that W ƒ and H ƒ are equal to the horizontal and vertical length of the LPF, respectively; generating a two-dimensional data array of the size W ƒ by H ƒ , wherein the data in each row of the array are all set to the corresponding pixel mean value of the same row in the W ƒ by H ƒ window of neighboring pixels; and applying the LPF to the array data to obtain the unsharp signal.

Claim 26

The method of claim 1 , further comprising the steps of, for a pixel that is not detected as belonging to an image edge in step (a): performing a filtering process on the pixel to obtain an unsharp signal at that pixel location; determining the difference between the original signal and the unsharp signal at the pixel location, wherein the difference represents image detail; boosting the image detail, and adding the boosted image detail to the original image signal to obtain a detail enhanced image.

Claim 27

An image detail enhancement system for enhancing an original image signal represented by a set of pixels, comprising: (a) a detector that detects image pixels that belong to an image edge; and (b) an edge enhancer that, for a detected pixel, generates an unsharp image signal at the detected pixel location, determines the difference between the original image signal and the unsharp signal, wherein said difference represents image details, boosts the difference signal and adds the boosted signal to the original signal to obtain a detail enhanced image signal, whereby boosting of any noise at the selected pixel location is reduced.

Claim 28

An image detail enhancement system for enhancing an original image signal represented by a set of pixels, comprising: (a) a detector that detects image pixels that belong to an image edge; and (b) an edge enhancer that, for a detected pixel, generates an unsharp image signal at the detected pixel location, determines the difference between the original image signal and the unsharp signal, wherein said difference represents image details, boosts the difference signal and adds the boosted signal to the original signal to obtain a detail enhanced image signal, whereby boosting of any noise at the selected pixel location is reduced.

Claim 29

The system of claim 26 wherein the detector further detects image pixels that belong to a horizontal or vertical image edge.

Claim 30

The system of claim 26 wherein the detector further detects image pixels that belong to a noisy and sharp horizontal or vertical image edge.

Claim 31

The system of claim 26 wherein the edge enhancer further generates an unsharp image signal by determining mean values of pixels in a two-dimensional window including the selected pixel, and using said mean values in a filtering process to obtain the unsharp image signal at the selected pixel location.

Claim 32

The system of claim 26 wherein in detecting image pixels that belong to an image edge, the detector selects a pixel and based on the selected pixel and its neighboring pixels in a two-dimensional window, determines if the pixel belongs to an image edge, wherein each pixel is checked separately to determine if it belongs to an image edge.

Claim 33

The system of claim 26 wherein if a pixel is determined to belong to an image edge, the detector further determining if the edge is noisy.

Claim 34

The system of claim 30 , wherein to detect image pixels that belong to an essentially vertical image edge, the detector further: selects at least W columns of pixels centered with the selected pixel, wherein each columns includes H pixels; determines the mean value m of the pixels in each column, thereby generating W mean values; determines the variance value σ of the pixels in each column, thereby generating W variance values; and based on the W mean values and the W variance values, determines if the selected pixel belongs to a vertical image edge.

Claim 35

The system of claim 32 , wherein: the mean values m j are determined according to the relation: m j = 1 H ⁢ ∑ i = - H - 1 2 H - 1 2 ⁢ I i , j , j = - 1 , 0 , 1 , the variance values σ j are determined according to the relation: σ j = 1 H ⁢ ∑ i = - H - 1 2 H - 1 2 ⁢  I i , j - m j  , j = - 1 , 0 , 1 , where I i,j is the value of a pixel located at row i and column j in the window, such that the row and column index of the selected pixel is 0.

Claim 36

The system of claim 32 , wherein in detecting if the selected pixel belongs to an image edge, the detector further: determines if |m 0 −m j |>max(σ 0 ,σ j ), j=−1 or 1; determines if max(|m 0 −m 1 |,|m 0 −m −1 |)≧T m ; where T m is a predetermined threshold value; such that if |m 0 −m j |>max(σ 0 ,σ j ), j=−1 or 1, and if max(|m 0 −m 1 |,|m 0 −m −1 |)≧T m , then the selected pixel is considered as a pixel in an image edge.

Claim 37

The system of claim 32 , wherein if the selected pixel is determined to belong to an image edge, the detector further determines if the edge is noisy.

Claim 38

The system of claim 35 , wherein in determining if the edge is noisy, the detector further: generates binary pattern data b from pixels in column 0 of the window according to the relation: b i , 0 = { 0 if ⁢ ⁢ I i , 0 < m 0 1 if ⁢ ⁢ I i , 0 ≥ m 0 ⁢ ⁢ i = - H - 1 2 , ⋯ ⁢ , 0 , ⋯ ⁢ , H - 1 2 ; based on the binary pattern data, generates a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ i = - H - 1 2 H - 1 2 - 1 ⁢  b i , 0 - b i + 1 , 0  ; compares the count N to a predetermined threshold value T N , 0<T N <H−1; wherein if the count N is not less than T N , then the edge is considered as noisy.

Claim 39

The system of claim 35 , wherein in determining if the edge is noisy, the detector further: generates binary pattern data b from pixels in column 0 of the window according to the relation: b i , 0 = { 0 if ⁢ ⁢ I i , 0 < m 0 1 if ⁢ ⁢ I i , 0 ≥ m 0 ⁢ ⁢ i = - H - 1 2 , ⋯ ⁢ , 0 , ⋯ ⁢ , H - 1 2 based on the binary pattern data, generates a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ i = - H - 1 2 H - 1 2 - 1 ⁢ ( b i , 0 ⊕ b i + 1 , 0 ) ; compares the count N to a predetermined threshold value T N , 0<T N <H−1; wherein if the count N is not less than T N , then the edge is considered as noisy.

Claim 40

The system of claim 29 , wherein: said window comprises a W ƒ by H ƒ rectangular window centered with the selected pixel; if the image edge direction is vertical, the edge enhancer further: determines the mean values of the pixels in each column of the window, wherein there are a total of W ƒ such mean values; includes a low pass filter (LPF) that filters the mean values to obtain an unsharp image signal at the selected pixel location.

Claim 41

The system of claim 38 , wherein the LPF comprises a one dimensional (1D) LPF, such that W ƒ is equal to the length of the LPF.

Claim 42

The system of claim 38 , wherein: the LPF comprises a two dimensional (2D) LPF, such that W ƒ and H ƒ are equal to the horizontal and vertical length of the LPF, respectively; the edge enhancer further generates a two-dimensional data array of the size W ƒ by H ƒ , wherein the data in each column of the array are all set to the corresponding pixel mean value of the same column in the W ƒ by H ƒ window of neighboring pixels; and the data array is filtered by the LPF to obtain the unsharp signal.

Claim 43

The system of claim 30 , wherein to detect image pixels that belong to an essentially horizontal image edge, the detector further: selects at least H rows of pixels centered with the selected pixel, wherein each row includes W pixels; determines the mean value of the pixels in each row, thereby generating H mean values; determines the variance value of the pixels in each row, thereby generating H variance values; based on the H mean values and the H variance values, determines if the selected pixel belongs to a horizontal image edge.

Claim 44

The system of claim 41 , wherein: the mean values m i are determined according to the relation: m i = 1 W ⁢ ∑ j = - W - 1 2 W - 1 2 ⁢ I i , j , i = - 1 , 0 , 1 , the variance values σ i are determined according to the relation: σ i = 1 W ⁢ ∑ j = - W - 1 2 W - 1 2 ⁢  I i , j - m i  , i = - 1 , 0 , 1 , where I i,j is the value of a pixel located at row i and column j in the window, such that the row and column index of the selected pixel is 0.

Claim 45

The system of claim 41 , wherein in detecting if the selected pixel belongs to an image edge, the detector further: determines if |m 0 −m i |>max(σ 0 ,σ i ), i=−1 or 1; determines if max(|m 0 −m 1 |,|m 0 −m −1 |)≧T m ; where T m is a predetermined threshold value; such that if |m 0 −m i |>max(σ 0 ,σ i ), i=−1 or 1, and if max(|m 0 −m 1 |,|m 0 −m −1 |)≧T m , then the selected pixel is considered as a pixel in an image edge.

Claim 46

The system of claim 41 , wherein if the selected pixel is determined to belong to an image edge, the detector further determines if the edge is noisy.

Claim 47

The system of claim 44 , wherein in detecting if the edge is noisy, the detector further: generates binary pattern data b from pixels in row 0 of the window according to the relation: b 0 , j = { 0 if ⁢ ⁢ I 0 , j < m 0 1 if ⁢ ⁢ I 0 , j ≥ m 0 ⁢ ⁢ j = - W - 1 2 , ⋯ ⁢ , 0 , ⋯ ⁢ , W - 1 2 ; based on the binary pattern data, generates a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ j = - W - 1 2 W - 1 2 - 1 ⁢  b 0 , j - b 0 , j + 1  ; compares the count N to a predetermined threshold value T N , 0<T N <W−1; wherein if the count N is not less than T N , then the edge is considered as noisy.

Claim 48

The system of claim 44 , wherein in detecting if the edge is noisy, the detector further: generates binary pattern data b from pixels in row 0 of the window according to the relation: b 0 , j = { 0 if ⁢ ⁢ I 0 , j < m 0 1 if ⁢ ⁢ I 0 , j ≥ m 0 ⁢ ⁢ j = - W - 1 2 , ⋯ ⁢ , 0 , ⋯ ⁢ , W - 1 2 ; based on the binary pattern data, generates a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ j = - W - 1 2 W - 1 2 - 1 ⁢ ( b 0 , j ⊕ b 0 , j + 1 ) ; compares the count N to a predetermined threshold value T N , 0<T N <W−1; wherein if the count N is not less than T N , then the edge is considered as noisy.

Claim 49

The system of claim 29 , wherein: said window comprises a W ƒ by H ƒ rectangular window centered with the selected pixel; if the image edge direction is horizontal, the edge enhancer further: determines the mean values of the pixels in each row of the window, wherein there are a total of H ƒ such mean values; includes a low pass filter (LPF) for performing a filtering process on the mean values to obtain an unsharp image signal at the selected pixel location.

Claim 50

The system of claim 47 , wherein the LPF comprises a one dimensional (1D) LPF, such that H ƒ is equal to the length of the LPF.

Claim 51

The system of claim 47 , wherein: the LPF comprises a two dimensional (2D) LPF, such that W ƒ and H ƒ are equal to the horizontal and vertical length of the LPF, respectively; the edge enhancer further generates a two-dimensional data array of the size W ƒ by H ƒ , wherein the data in each row of the array are all set to the corresponding pixel mean value of the same row in the W ƒ by H ƒ window of neighboring pixels; and LPF filters the array data to obtain the unsharp signal.

Claim 52

The system of claim 26 , further comprising a normal enhancer, wherein for a pixel that is not detected as belonging to an image edge, the normal enhancer performs a filtering process on the pixel to obtain an unsharp signal at that pixel location, determines the difference between the original signal and the unsharp signal at the pixel location, wherein the difference represents image detail, boosts the image detail and adds the boosted image detail to the original image signal to obtain a detail enhanced image.

Claim 53

An image detail enhancement system for enhancing an original image signal represented by a set of pixels, comprising: (a) a detector that detects image pixels that belong to an image edge; (b) an edge enhancer that, for a detected pixel, generates an unsharp image signal at the detected pixel location, determines the difference between the original image signal and the unsharp signal, wherein said difference represents image details, boosts the difference signal and adds the boosted signal to the original signal to obtain a detail enhanced image signal, whereby boosting of any noise at the selected pixel location is reduced; and (c) a normal enhancer, wherein for a pixel that is not detected as belonging to an image edge, the normal enhancer performs a filtering process on the pixel to obtain an unsharp signal at that pixel location, determines the difference between the original signal and the unsharp signal at the pixel location, wherein the difference represents image detail, boosts the image detail and adds the boosted image detail to the original image signal to obtain a detail enhanced image; wherein upon determining that an image pixel belongs to an image edge, the detector selects the edge enhancer to enhance that image pixel, and upon determining that an image pixel does not belong to an image edge, the detector selects the normal enhancer to enhance that image pixel.

Claim 54

An image detail enhancement system for enhancing an original image signal represented by a set of pixels, comprising: (a) a detector that detects image pixels that belong to an image edge; (b) an edge enhancer that, for a detected pixel, generates an unsharp image signal at the detected pixel location, determines the difference between the original image signal and the unsharp signal, wherein said difference represents image details, boosts the difference signal and adds the boosted signal to the original signal to obtain a detail enhanced image signal, whereby boosting of any noise at the selected pixel location is reduced; and (c) a normal enhancer, wherein for a pixel that is not detected as belonging to an image edge, the normal enhancer performs a filtering process on the pixel to obtain an unsharp signal at that pixel location, determines the difference between the original signal and the unsharp signal at the pixel location, wherein the difference represents image detail, boosts the image detail and adds the boosted image detail to the original image signal to obtain a detail enhanced image; wherein upon determining that an image pixel belongs to an image edge, the detector selects the edge enhancer to enhance that image pixel, and upon determining that an image pixel does not belong to an image edge, the detector selects the normal enhancer to enhance that image pixel.

Claim 55

The system of claim 51 wherein the detector further detects image pixels that belong to a horizontal or vertical image edge.

Claim 56

The system of claim 51 wherein the detector further detects image pixels that belong to a noisy and sharp horizontal or vertical image edge.

Claim 57

The system of claim 51 wherein the edge enhancer further generates an unsharp image signal by determining mean values of pixels in a two-dimensional window including the selected pixel, and using said mean values in a filtering process to obtain the unsharp image signal at the selected pixel location.

Claim 58

The system of claim 51 wherein in detecting image pixels that belong to an image edge, the detector selects a pixel and based on the selected pixel and its neighboring pixels in a two-dimensional window, determines if the pixel belongs to an image edge, wherein each pixel is checked separately to determine if it belongs to an image edge.

Claim 59

The system of claim 51 wherein if a pixel is determined to belong to an image edge, the detector further determining if the edge is noisy.

Claim 60

The system of claim 55 , wherein to detect image pixels that belong to an essentially vertical image edge, the detector further: selects at least W columns of pixels centered with the selected pixel, wherein each columns includes H pixels; determines the mean value m of the pixels in each column, thereby generating W mean values; determines the variance value a of the pixels in each column, thereby generating W variance values; and based on the W mean values and the W variance values, determines if the selected pixel belongs to a vertical image edge.

Claim 61

The system of claim 57 , wherein: the mean values m j are determined according to the relation: m j = 1 H ⁢ ∑ i = - H - 1 2 H - 1 2 ⁢ I i , j , j = - 1 , 0 , 1 , the variance values σ j are determined according to the relation: σ j = 1 H ⁢ ∑ i = - H - 1 2 H - 1 2 ⁢  I i , j - m j  , j = - 1 , 0 , 1 , where I i,j is the value of a pixel located at row i and column j in the window, such that the row and column index of the selected pixel is 0.

Claim 62

The system of claim 57 , wherein in detecting if the selected pixel belongs to an image edge, the detector further: determines if |m 0 −m j |>max(σ 0 ,σ j ), j=−1 or 1; determines if max(|m 0 −m 1 |,|m 0 −m −1 |)≧T m ; where T m is a predetermined threshold value; such that if |m 0 −m j |>max(σ 0 ,σ j ), j=−1 or 1, and if max(|m 0 −m 1 |,|m 0 −m −1 |)≧T m , then the selected pixel is considered as a pixel in an image edge.

Claim 63

The system of claim 57 , wherein if the selected pixel is determined to belong to an image edge, the detector further determines if the edge is noisy.

Claim 64

The system of claim 60 , wherein in determining if the edge is noisy, the detector further: generates binary pattern data b from pixels in column 0 of the window according to the relation: b i , 0 = { 0 if ⁢ ⁢ I i , 0 < m 0 1 if ⁢ ⁢ I i , 0 ≥ m 0 ⁢ ⁢ i = - H - 1 2 , ⋯ ⁢ , 0 , ⋯ ⁢ , H - 1 2 ; based on the binary pattern data, generates a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ i = - H - 1 2 H - 1 2 - 1 ⁢  b i , 0 - b i + 1 , 0  ; compares the count N to a predetermined threshold value T N , 0<T N <H−1; wherein if the count N is not less than T N , then the edge is considered as noisy.

Claim 65

The system of claim 60 , wherein in determining if the edge is noisy, the detector further: generates binary pattern data b from pixels in column 0 of the window according to the relation: b i , 0 = { 0 if ⁢ ⁢ I i , 0 < m 0 1 if ⁢ ⁢ I i , 0 ≥ m 0 ⁢ ⁢ i = - H - 1 2 , ⋯ ⁢ , 0 , ⋯ ⁢ , H - 1 2 ; based on the binary pattern data, generates a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ i = - H - 1 2 H - 1 2 - 1 ⁢ ( b i , 0 ⊕ b i + 1 , 0 ) ; compares the count N to a predetermined threshold value T N , 0<T N <H−1; wherein if the count N is not less than T N , then the edge is considered as noisy.

Claim 66

The system of claim 54 , wherein: said window comprises a W ƒ by H ƒ rectangular window centered with the selected pixel; the edge enhancer includes a vertical edge enhancer, such that if the image edge direction is vertical, the detector selects the vertical edge enhancer such that the vertical edge enhancer: determines the mean values of the pixels in each column of the window, wherein there are a total of W ƒ such mean values; and includes a low pass filter (LPF) that filters the mean values to obtain an unsharp image signal at the selected pixel location.

Claim 67

The system of claim 63 , wherein the LPF comprises a one dimensional (1D) LPF, such that W ƒ is equal to the length of the LPF.

Claim 68

The system of claim 63 , wherein: the LPF comprises a two dimensional (2D) LPF, such that W ƒ and H ƒ are equal to the horizontal and vertical length of the LPF, respectively; the vertical edge enhancer further generates a two-dimensional data array of the size W ƒ by H ƒ , wherein the data in each column of the array are all set to the corresponding pixel mean value of the same column in the W ƒ by H ƒ window of neighboring pixels; and the data array is filtered by the LPF to obtain the unsharp signal.

Claim 69

The system of claim 55 , wherein to detect image pixels that belong to an essentially horizontal image edge, the detector further: selects at least H rows of pixels centered with the selected pixel, wherein each row includes W pixels; determines the mean value of the pixels in each row, thereby generating H mean values; determines the variance value of the pixels in each row, thereby generating H variance values; based on the H mean values and the H variance values, determines if the selected pixel belongs to a horizontal image edge.

Claim 70

The system of claim 66 , wherein: the mean values m i are determined according to the relation: m i = 1 W ⁢ ∑ j = - W - 1 2 W - 1 2 ⁢ I i , j , i = - 1 , 0 , 1 , the variance values σ i are determined according to the relation: σ i = 1 W ⁢ ∑ j = - W - 1 2 W - 1 2 ⁢  I i , j - m i  , i = - 1 , 0 , 1 , where I i,j is the value of a pixel located at row i and column j in the window, such that the row and column index of the selected pixel is 0.

Claim 71

The system of claim 66 , wherein in detecting if the selected pixel belongs to an image edge, the detector further: determines if |m 0 −m i |>max(σ 0 ,σ i ), i=−1 or 1; determines if max(|m 0 −m 1 |,|m 0 −m −1 |)≧T m ; where T m is a predetermined threshold value; such that if |m 0 −m i |>max(σ 0 ,σ i ), i=−1 or 1, and if max(|m 0 −m 1 ,|m 0 −m −1 |)≧T m , then the selected pixel is considered as a pixel in an image edge.

Claim 72

The system of claim 66 , wherein if the selected pixel is determined to belong to an image edge, the detector further determines if the edge is noisy.

Claim 73

The system of claim 69 , wherein in detecting if the edge is noisy, the detector further: generates binary pattern data b from pixels in row 0 of the window according to the relation: b 0 , j = { 0 if ⁢ ⁢ I 0 , j < m 0 1 if ⁢ ⁢ I 0 , j ≥ m 0 j = - W - 1 2 , ⋯ ⁢ , 0 , ⋯ ⁢ , W - 1 2 ; based on the binary pattern data, generates a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ j = - W - 1 2 W - 1 2 - 1 ⁢  b 0 , j - b 0 , j + 1  ; compares the count N to a predetermined threshold value T N , 0<T N <W−1; wherein if the count N is not less than T N , then the edge is considered as noisy.

Claim 74

The system of claim 69 , wherein in detecting if the edge is noisy, the detector further: generates binary pattern data b from pixels in row 0 of the window according to the relation: b 0 , j = { 0 if ⁢ ⁢ I 0 , j < m 0 1 if ⁢ ⁢ I 0 , j ≥ m 0 j = - W - 1 2 , ⋯ ⁢ , 0 , ⋯ ⁢ , W - 1 2 ; based on the binary pattern data, generates a count N of the number of neighboring binary pattern data that vary from each other according to the relation: N = ∑ j = - W - 1 2 W - 1 2 - 1 ⁢ ( b 0 , j ⊕ b 0 , j + 1 ) ; compares the count N to a predetermined threshold value T N , 0<T N <W−1; wherein if the count N is not less than T N , then the edge is considered as noisy.

Claim 75

The system of claim 54 , wherein: said window comprises a W ƒ by H ƒ rectangular window centered with the selected pixel; the edge enhancer includes a horizontal edge enhancer, wherein if the image edge direction is horizontal, the detector selects the horizontal edge enhancer such that the horizontal edge enhancer: determines the mean values of the pixels in each row of the window, wherein there are a total of H ƒ such mean values; and includes a low pass filter (LPF) for performing a filtering process on the mean values to obtain an unsharp image signal at the selected pixel location.

Claim 76

The system of claim 72 , wherein the LPF comprises a one dimensional (1D) LPF, such that H ƒ is equal to the length of the LPF.

Claim 77

The system of claim 72 , wherein: the LPF comprises a two dimensional (2D) LPF, such that W ƒ and H ƒ are equal to the horizontal and vertical length of the LPF, respectively; the horizontal edge enhancer further generates a two-dimensional data array of the size W ƒ by H ƒ , wherein the data in each row of the array are all set to the corresponding pixel mean value of the same row in the W ƒ by H ƒ window of neighboring pixels; and LPF filters the array data to obtain the unsharp signal.