JP3193543B2 - Video signal correction device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is intended to correct color fringing in video equipment handling color images, video equipment such as full color printers, full color faxes, and information equipment capable of storing images using magnetic disks and optical disks. The present invention relates to a video signal correction device for improving the image quality.

[0002]

2. Description of the Related Art In recent years, with the development of full-color hard copy technology, a faithful image can be reproduced using a printing technology such as a sublimation type thermal transfer system, and the demand for a still image output is increasing. In color reproduction as well, high-definition video signals such as high-definition video signals have the same performance as silver halide photography.

However, the current system (such as the NTSC system)
Is output as a still image, the resolution is limited by the band limitation of the video signal. In particular, the color difference signal can only obtain a resolution of about 1/3 or less as compared with the luminance signal. An image whose color should change as the luminance changes cannot be followed by a change in the color difference value, and is an image recorded at a portion after the luminance change using the color before the luminance change. That is, an image with a lot of color blur will be recorded.

Therefore, as a method of removing the color fringing, for example, a method has been proposed in which contour information in a luminance signal is added to a color difference signal to correct the color difference signal (Japanese Patent Laid-Open No. 2-21).
No. 3282).

FIG. 29 is a block diagram showing a conventional video signal correction apparatus for correcting color fringing.
In FIG. 29, the luminance signal Y and the color difference signals RY and BY output from the digital signal source 100 are connected to the connection device 1.
01 and stored in the image memory 102. The contour correction circuit 103 performs both horizontal and vertical aperture corrections by digital processing based on the luminance signal read from the image memory 102.

After that, the luminance signal whose outline is corrected by the outline correction circuit 103 is output to the digital signal processor 10.
6 and the contour correction component is output to the addition circuit 105 as edge information. On the other hand, the color difference signals RY and BY fetched from the image memory 102 have their noise components reduced by a low-pass filter (LPF) 104, and then are subjected to contour correction for the above-mentioned luminance signal by adding circuits 105A and 105B, respectively. The edge information of the components is added and output to the digital signal processing unit 106.

The digital signal processing section 106 performs chroma enhancement, γ correction, and the like based on the input luminance signal and color difference signals RY and BY, and outputs a digital signal to the printer 109. Further, when one frame is composed of two fields, such as an NTSC signal, movement between the connection device 101 and the image memory 102 is performed to eliminate the occurrence of image blur in the first and second fields. A detection circuit 107 is provided.
7, a scanning line interpolation circuit 108 for obtaining a frame image from a field is connected. Also, the printer 10
Numeral 9 prints out an image on an appropriate sheet of paper based on an output signal from the digital signal processing unit 106.

[0008]

However, in the above configuration, when the band of the color difference signal is extremely narrower than the luminance signal like the NTSC signal, the contour correction component of the luminance signal is added to the color difference signal. However, blurring of the color difference signal is not corrected well, and conversely, a portion having a different color occurs in the contour correction portion of the luminance signal, thereby further deteriorating the image quality. Further, since the contour correction component is added even in a place where the value of the color difference signal is constant, there is a problem that a pseudo contour is easily generated.

SUMMARY OF THE INVENTION The present invention provides a video signal correction apparatus which can improve the degree of correction of color bleeding of an image and suppress image quality deterioration at the time of correction in consideration of such a problem in the conventional correction of color bleeding. It is intended for.

[0010]

According to the present invention, a storage means for storing a luminance value and a chrominance value of each pixel of an image, and a luminance signal generated from the stored luminance value are inputted, and a predetermined signal of the image is inputted. A brightness change position detecting means for detecting a position of a pixel at at least one end of a brightness change region in which a brightness value in the direction increases or decreases by a predetermined value or more, and from the detected position of the pixel at the end to the outside of the brightness change region. A luminance stable area detecting means for detecting a luminance stable area in which a change in the luminance value is equal to or less than a predetermined luminance value, and a color difference value in a predetermined direction of the image within the detected luminance stable area increases by a predetermined value or more. Alternatively, a color difference value change area that decreases is detected, and at least a part of the color difference values stored in the storage means in the detected color difference value change area is converted to a color difference value stable area with little change in the color difference value outside the color difference value change area. of With difference values, the color difference value correction means for correcting, continuously in a predetermined direction perpendicular to the direction of the image in which the color difference value is corrected, the luminance value falls within a predetermined value,
A video signal correction apparatus including a vertical color difference value correction unit that selects a plurality of pixels and smoothes the color difference values of those pixels.

[0011]

According to the present invention, the storage means stores a luminance value and a color difference value of each pixel of an image, and the luminance change position detecting means detects a luminance change area in which the luminance value in a predetermined direction of the image increases or decreases by a predetermined value or more. The luminance stable area detecting means detects a luminance stable area in which a change in luminance value is equal to or less than a predetermined luminance value from the position of the pixel at the end to the outside of the luminance change area. The color difference value correction means detects the color difference value change area in the luminance stable area where the color difference value in the predetermined direction of the image increases or decreases by a predetermined value or more, and stores the color difference value change area in the storage means in the color difference value change area. The corrected at least a part of the color difference values are corrected by using the color difference values of the color difference value stable region with little change in the color difference value, and the vertical color difference value correction unit is configured to correct the color difference value in a direction perpendicular to a predetermined direction of the image in which the color difference values are corrected The brightness value Fall within a range of values, by selecting the respective plurality of picture, even out the color difference values of those pixels. That is, by correcting the color difference signal so that the change in the color difference value matches the change in the luminance value, it is possible to correct the dullness of the color difference signal, reduce color fringing, and improve image quality.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the drawings showing embodiments thereof.

FIG. 1 is a block diagram of a video signal correction device according to a first embodiment of the present invention. In FIG. 1, 1
Reference numeral denotes an image memory which is storage means for sequentially storing a luminance value and two color difference values of a pixel continuous in a horizontal scanning direction (hereinafter, referred to as a horizontal direction) of a video signal so as to correspond to a pixel position.

Reference numeral 2 denotes an edge start point detecting means for detecting an n1st pixel whose luminance value starts to increase or decrease continuously in two or more pixels in the horizontal direction as an edge start point, and reference numeral 3 denotes an edge start point detecting means 2. When the increase or decrease of the luminance value continuous in the horizontal direction of the image from the edge start point ends n
Edge end point detecting means for detecting the second pixel as an edge end point.

Reference numeral 4 denotes a k1-th pixel between the (n1-a) th where the variation of the luminance value is small and the edge start point, and the k1-th pixel whose chrominance value starts to increase or decrease continuously from the edge start point. The edge start point color difference value correction means changes at least a part of the color difference values of the pixels from the pixel to the edge start point.

Reference numeral 5 denotes at least a part of the color difference values of the pixels from the edge end point where the variation of the luminance value is small to the (n2 + b) th pixel and from the edge end point to the k2th pixel where the color difference value continuously increases or decreases. Is a color difference value correction unit for changing the edge end point.

6 reads out the luminance and chrominance values of at least three consecutive pixels in the vertical direction from the image memory 1 in which the chrominance values corrected in the horizontal direction are stored, and assigns one pixel to the pixel of interest. And the luminance value Yh of the pixel of interest
A vertical chrominance value correction unit that selects a pixel whose absolute value of the luminance value difference is within δ from the remaining pixels, and changes the chrominance value of the target pixel using the chrominance value of the selected pixel. .

Next, the operation of the above embodiment will be described.

First, while the luminance value of the image is taken out from the image memory 1 in the horizontal direction, the edge start point detecting means 2 starts to change the luminance value greatly. The edge end point which is the n2th pixel at which the value change ends is detected. Next, while extracting color difference values of the image in the horizontal direction from the image memory 1, if the color difference values continuously change immediately before the edge start point or immediately after the edge end point, the edge start point portion color difference value correction means 4 and the edge end point The color difference value correcting means 5 changes the color difference value of the area where the color difference values are changing.

Next, a luminance value of a total of three pixels is extracted from the image memory 1 including one pixel of interest and pixels located above and below the pixel of interest, and the luminance value Yh of the pixel of interest is calculated by the vertical chrominance value correction means 6. As a reference, a pixel whose absolute value of the difference from the luminance value of the remaining pixels is within δ is selected, and the color difference value of the target pixel is corrected and output using the color difference value of the selected pixel. In this way, the color difference values of the pixels that are continuous in the vertical direction are corrected, and the color difference values are transferred to the image memory 1 when the correction is completed.

Here, the storing means is realized by the image memory 1, the luminance change position detecting means is constituted by the edge starting point detecting means 2 and the edge ending point detecting means 3, and the luminance stable area detecting means and the color difference value correcting means are: An edge start point color difference value correction unit 4 and an edge end point color difference value correction unit 5 are provided.

The above processing is performed on all the image data, and a processed image with reduced color fringing can be obtained.

The image memory 1 continuously reads, for example, pixel values (luminance value and color difference value) of pixels that are continuous in the horizontal direction of the image in order to facilitate reading of luminance values, reading and changing of color difference values. It is preferable to use those that can. However, when the image memory 1 is used that cannot read the pixel values of the pixels that are continuous in the horizontal direction of the image, the pixel values that are continuous in the horizontal direction are temporarily read while reading the pixel values from the image memory 1. A configuration in which a line memory or the like for storing is further used may be adopted.

The image memory 1 is preferably a page memory capable of storing at least one image. However, even if the image memory 1 is constituted by a plurality of line memories, for example, a capacity capable of temporarily storing an image signal while each processing is completed. Good.

The detection of the continuous increase / decrease of the brightness value in the edge start point detection means 2 and the edge end point detection means 3 is, for example, two or more cases where the sign of the difference value of the brightness value of the adjacent pixels is the same, positive or negative. Detecting a continuous case, or two or more consecutive cases where the sign of the difference value of the luminance value of adjacent pixels is the same as positive or negative and the case where the sum of the difference values is greater than or equal to a preset value , Or by detecting a case where two or more cases where the difference value of the luminance value between adjacent pixels is equal to or more than a predetermined threshold value are continuous.

Since a general image signal, especially an NTSC signal, has a lot of noise, when detecting a continuous increase or decrease in the luminance value, a large number of small noise signals are detected, and the processing time tends to be long. In view of this, it is necessary to confirm the continuous increase / decrease of the brightness value by continuing three or more pixels instead of two pixels, or to slightly increase the set value for the sum of the difference values or the threshold value for each difference value by the method described above. preferable.

The increase or decrease of the luminance value can be made proportional to the number of pixels whose increase and decrease are continuous depending on the difference value between adjacent pixels. In this case, the color blur can be reduced not only when the luminance value changes steeply as shown in this embodiment, but also when the luminance value changes gradually.

The edge start point chrominance value correction means 4 recognizes the (n1-a) th change in the luminance value which is small, for example, while extracting the luminance value from the edge start point in reverse order, the absolute value of the difference value is larger than a predetermined threshold value. Pixel is n
A pixel which is determined to be the 1-a-th pixel or a pixel whose difference in luminance value from the luminance value of the edge start point does not fall within a predetermined constant α or Z% of the luminance value of the edge start point is n1-a.
This can be performed by setting the pixel at the n-th pixel or the pixel at the end point of another edge immediately before the start point of the edge currently searched for as the (n1-a) th pixel.

It should be noted that n1-a having a small variation in the luminance value.
The determination of the region from the first to the edge start point can be performed by either the edge start point detection means 2 or the edge start point color difference value correction means 4.

The edge end point chrominance value correcting means 5 also recognizes the (n2 + b) th from the edge end point where the variation of the luminance value is small, similarly to the edge start point chrominance value correcting means 4.
For example, a pixel in which the absolute value of the difference value is larger than a predetermined threshold value is taken as the n2 + b-th pixel while successive luminance values are sequentially taken out from the edge end point, or the difference between the luminance value and the luminance value of the edge end point is predetermined. A pixel that does not fall within the constant α or Z% of the luminance value of the edge end point is set to the n2 + b-th pixel, or the pixel of the edge start point that appears immediately after the edge end point currently being searched is set to n.
This can be performed by setting the 2 + b-th pixel.

The determination of the area from the edge end point where the change in the luminance value is small to the n2 + b-th area can be performed by either the edge end point detecting means 3 or the edge end point color difference value correcting means 5.

The detection of the continuous increase / decrease of the color difference values in the edge start point color difference value correction means 4 and the edge end point color difference value correction means 5 is performed, for example, when the sign of the difference value of the color difference value between adjacent pixels is positive or negative and the same. If two or more consecutive cases are detected, two or more consecutive cases where the sign of the difference value of the color difference value of the adjacent pixel is the same as positive or negative and the sum of the difference values are set in advance It can be executed by detecting a case where the difference value is equal to or more than the set value or detecting a case where two or more cases where the difference value of the color difference value between adjacent pixels is equal to or more than a preset threshold value are consecutive.

Further, in the case where the band of the color difference signal is extremely narrow as compared with the luminance signal, for example, in the case of an NTSC signal or the like, to detect the increase or decrease of the color difference value by the above-mentioned method, the fluctuation of the color difference value of the adjacent pixel is small. Therefore, it is preferable to perform the processing on the difference value of the color difference value between pixels sandwiching one or more pixels. Furthermore, since a general image signal, particularly an NTSC signal, includes noise, when detecting a continuous increase or decrease in the color difference value, a large number of small noise signals are detected, and the processing time tends to be long. Therefore, in order to perform the processing in a short time, it is confirmed that the increase / decrease of the color difference value is continuous not less than 2 pixels but 3 pixels or more. It is preferable to slightly increase the threshold value.

The change of the color difference value in the edge start point color difference value correction means 4 uses the color difference value in the region from the (n1-a) th pixel to the edge start point in which the color difference value does not vary much. For example, k1 to k1-c (≧≧) with little change in color difference value
The color difference value is changed to the representative value C1 of the color difference value of each of the (n1-a) th pixels, or the average value of the representative value C1 and each color difference value is changed.

The change of the color difference value in the edge end point color difference value correcting means 5 uses a color difference value in a region from the edge end point to the n2 + b-th pixel where the color difference value is small. For example, k2 to k2 + d (≦≦
The color difference value is changed to the representative value C2 of the color difference value of each (n2 + b) -th pixel, or the average value of the representative value C2 and each color difference value is changed.

The representative value C1 and the representative value C2 are average values used for general frequency distribution such as an additive average value, a multiplication average, a harmonic average value, a median value, and a mode value of the color difference values of a plurality of selected pixels. Can be used. In this case, if only one pixel is selected, it is preferable not to change the color difference value.

The vertical chrominance value correcting means 6 must refer to the chrominance value before being processed when it is referred to as the process for the next adjacent pixel.
It also has a function of temporarily storing the color difference values of already processed pixels so as not to write them to the image memory 1. This can be configured by, for example, a line memory that temporarily stores the processing results in the vertical direction, or a plurality of buffers having a capacity that can be stored while the pixel position for storing the processing results in the vertical direction is within the processing range. .

In this embodiment, the vertical chrominance value correcting means 6 corrects the chrominance value with reference to three vertically consecutive pixels (including the target pixel).
Pixels and seven pixels can be referred to. However, in this case, not only a pixel whose luminance value is at least close to the luminance value of the target pixel but also only a pixel that is continuous with the target pixel by the connection of pixels having similar luminance values is selected.

The correction of the color difference value by the vertical color difference value correction means 6 is performed once or more for the same pixel. by the way,
The color noise originally occurring in the horizontal direction and the remainder of the color fringing portion generated when the color difference value correction in the horizontal direction in the color fringing portion is not completely processed are converted into vertical color difference values by the vertical color difference value correcting means 6. It can be diffused in the direction and cannot be recognized. The effect is so great that the processing in the vertical direction for the same pixel is repeated, and it is preferable to repeat the processing three or more times.

The correction of the chrominance value by the vertical chrominance value correction means 6 selects a pixel having a similar difference between the luminance value of the pixel of interest and the luminance values of the pixels above and below it, and selects the chrominance value of the selected pixel and the h-th pixel. The luminance value is replaced by the average value, but the h-th new color difference value may be obtained by adding or subtracting the Z% of the difference from the color difference value of the selected pixel to the h-th color difference value. it can.

By performing the above-described color difference value correction on the entire image data by the vertical color difference value correction means 6, color noise slightly generated by the color bleeding process can be reduced, and a processed image of good image quality can be obtained. it can.

FIG. 2 is a block diagram of a video signal correction device in which the embodiment shown in FIG. 1 is realized by using a microcomputer.

In FIG. 2, reference numeral 7 denotes a luminance value Y, a color difference value R-Y, and a luminance value Y continuously obtained in the horizontal direction of the input video signal.
An external interface for converting the color difference value BY into an 8-bit digital signal. The digital interface 8 stores and processes the luminance value Y, color difference value R-Y, and color difference value BY converted into the digital signal. It is an image memory for storing the resulting color difference values.

The image memory 8 has a storage capacity for three pages for storing each value. The luminance value and two color difference values that make up one pixel in the memory are obtained by vertical and horizontal addresses common to each page. Page 0 stores the luminance value Y, and page 1 stores the color difference values RY,
Page 2 stores color difference values BY.

Reference numeral 9 denotes an edge start point detection means 2, an edge end point detection means 3, an edge start point color difference value correction means 4, an edge end point color difference value correction means 5, and a vertical color difference value correction means 9 for an input video signal. 6 is a microcomputer that performs each process of FIG. Note that the microcomputer 9 shown here has a CPU, a ROM, a RAM, and an input / output unit. Reference numeral 10 denotes an image line memory used as a temporary storage auxiliary memory for storing one line of values obtained by correcting the color difference values by the vertical color difference value correcting means 6 in the vertical direction.

Next, an example of the operation of the video signal processing device configured as described above will be described with reference to FIGS.

FIG. 3 shows the relationship between the storage position of the image memory 8 and the position of the pixel used in this embodiment. In this case, the horizontal scanning direction of the video signal matches the horizontal direction (x direction) of the image memory 8, and the sub-scanning direction matches the vertical direction (y direction) of the image memory. (X, y) represents a pixel position in the image, and corresponds to each address of the image memory 8.

FIG. 4 is a flowchart showing the overall processing of the video signal correction device. In FIG. 4, first, the color difference value is corrected in the horizontal direction of the image to reduce color fringing (step s1), and then processing in the vertical direction of the image is performed (step s2) to reduce color noise and reduce the color noise. Improve image quality for color. That is, in step s1,
Correction of the color difference value in the horizontal direction is realized, and correction of the color difference value in the vertical direction is realized in step s2.

Hereinafter, the correction of the color difference values in the horizontal direction realized by the processing of step s1 will be described with reference to FIGS. 5 to 9, and the correction of the color difference values in the vertical direction realized by the processing of step s2 will be described with reference to FIG. This will be described with reference to FIG.

FIG. 5 is a flowchart showing the entire process of correcting color difference values in the horizontal direction. Each process will be described below.

In step s3, the position of a pixel in the image memory 8 at which processing in the vertical direction is started is set.

In step s4, the position of a pixel in the image memory 8 at which horizontal processing is started is set.

In step s5, a page 0 for extracting a luminance value from the image memory 8 is set.

In step s6, the luminance values are sequentially extracted from the image memory 8, and the pixel position n1 of the edge start point of the luminance value is obtained.
Then, an area where the luminance value does not fluctuate in the pixels before the edge start point is obtained (corresponding to the processing of the edge start point detection means 2 in FIG. 1).

In step s7, the detected edge start point 3
It is checked whether the pixel ahead of the pixel does not exceed the horizontal end point of the image. If the horizontal end point is exceeded, the edge end point cannot be detected, so that the color difference value is not corrected.

In step s8, the luminance value is taken out from the image memory 8, the edge end point of the luminance value is found, and the position information n2 and the area where the luminance value does not fluctuate in the pixels after the edge end point are obtained (the edge in FIG. 1). This corresponds to the processing of the end point detecting means 3).

In step s9, the setting of the page of the image memory 8 is switched to read the value of the color difference value (setting for extracting any one of the color difference values RY and BY) is performed.

In step s10, the color difference values of the pixels in the region where the luminance value obtained in step s6 is small are sequentially extracted from the storage position of the image memory 8, the k1st pixel is detected, and the edge from the k1th pixel is detected. The color difference value up to the start point is corrected and output to the image memory 8 (corresponding to the processing of the edge start point color difference value correction means 4 in FIG. 1).

In step s11, the color difference values of the pixels in the region where the luminance value obtained in step s8 is small are sequentially taken out from the storage position of the image memory 8, the k2th pixel is detected, and the edge end point to the k2th pixel are detected. Is corrected and output to the image memory 8 (corresponding to the processing of the color difference value correcting means 5 at the edge end point in FIG. 1).

In step s12, each color difference value RY, B-
It is determined whether to switch the page of the image memory for performing the color difference value correction for Y.

In step s13, the pixel position to be processed is
It is determined whether the processing for one line in the horizontal direction of the image has been completed. If there is an unprocessed portion, the process from step s5 is repeated to correct the color difference value for all pixels in the horizontal direction.

In step s14, if the processing for one line in the horizontal direction has been completed in step s13, it is determined whether or not the video signal correction processing has been completed for the entire image to be processed. If the processing has not been completed, step s15
, The address y in the vertical direction is increased by one, and the above-described processing after step s4 is repeated.

Next, the aforementioned steps s6, s8, s1
The processing of 0 and s11 will be described in detail with reference to FIGS. 6, 7, 8, and 9.

FIG. 6 shows the processing for obtaining the edge start point n1 in step s6 in FIG. In FIG.
In step s20, a variable c for counting the number of pixels whose luminance value continuously increases or decreases and a luminance value of the edge start point n1 as Y1 and Y1 to α or more and Y1 + α or less (α is previously determined to be an appropriate value. Initialize a variable a (that is, the number of pixels for which the luminance value is considered to be stable) for counting the number of continuous pixels having the luminance value of (a).

In step s21, the difference between the luminance value of the position x of the pixel to be processed and the adjacent pixel x + 1 is calculated, and
The pixel whose absolute value of the difference value is equal to or larger than the threshold value Th1 at 22 is defined as a pixel whose luminance value increases or decreases. If the condition is not satisfied, the pixel position is shifted by one in step s23, and the process is repeated until the condition in step s22 is satisfied.

If it is determined in step s22 that the pixel is to increase or decrease, the sign state of the difference value is stored in a flag in step s24. The flag is a flag for knowing whether the luminance value is continuously increasing or decreasing.

Next, in step s25, the pixel position to be processed is shifted by one, and in step s26, the difference between the position x of the pixel to be processed and the luminance value of the adjacent pixel x + 1 is calculated, and step s27 is performed.
It is determined whether or not the pixel increases or decreases as in step s22. If the condition is not satisfied, step s
At step 28, the number c of pixels that continuously increase or decrease is returned to the initial value “0”, the pixel position to be processed is shifted by one at step s23, and the process returns to step s21. Then, the process of counting the number of pixels that continuously increase or decrease is repeated.

If the condition is satisfied in step s27,
In step s29, the sign state of the difference value is determined, and the sign state and the sign state of the flag are compared.

If the sign is different as a result of the comparison, it is determined that the luminance value has not continuously increased or decreased, the flag is changed to the sign obtained in step s29 in step s30, and the flag is continuously increased in step s31. Alternatively, the number of decreasing pixels c is initialized to “0”, and the process returns to step s25 and is repeated.

On the other hand, if the sign is the same sign as a result of the comparison, it is determined that the luminance value is continuously increasing or decreasing, and a variable for counting the number of pixels continuously increasing or decreasing is determined in step s32. Increment c by one.

Next, in step s33, if the value of the number of pixels c continuously increasing or decreasing is 3 or more, it is determined that there are at least three pixels in which the luminance value continuously increases or decreases, and the next Move on to processing. Conversely, if the value of the number of pixels c is less than 3, the process returns to step s25. That is, in the present embodiment, in consideration of the processing time, the continuous increase or decrease of the luminance value is detected when three pixels continue.

If the value of the number of pixels c is 3 or more, step s
At 34, the pixel position (x-2) is substituted for the edge start point n1. Next, it is confirmed whether there is a number of pixels adjacent to this pixel whose luminance is considered to be stable.

In step s35, the luminance value Y1 of the edge start point is extracted.

In step s36, the pixel position to be processed is set to a pixel position adjacent to the edge start point.

At step s37, the luminance values of the pixels at the positions continuing from the edge start point are sequentially extracted.

In step s38, it is determined whether or not the extracted luminance value is equal to or more than Y1-α and equal to or less than Y1 + α.

As a result of the judgment, if the luminance value is within the range, the number of pixels a satisfying the condition of step s38 in step s39
Is counted by one, the pixel position to be processed is shifted by one in step s40, and the process returns to step s37 to obtain the number of pixels a for which the luminance value is considered to be stable.

On the other hand, if the extracted luminance value falls outside the luminance value range from Y1-α to Y1 + α in step s38, the processing is terminated and the edge start point n1 and the end of the region where the fluctuation of the luminance value is considered to be small are determined. A certain n1-ath pixel position is determined.

FIG. 7 shows the processing for obtaining the edge end point n2 in step s8 in FIG. In FIG.
In step s50, the luminance values Y2 to Y2-
β or more and Y2 + β or less (β is determined to an appropriate value in advance)
The variable b for counting the number of continuous pixels having the luminance value (that is, the number of pixels whose luminance value is considered to be stable) is initialized.

In step s51, the difference between the luminance value of the input pixel x at the position of the (n1 + 2) th pixel and the luminance value of the adjacent pixel x + 1 is obtained, and in step s52, the sign state of the difference value is stored in the flag. The flag is a flag for knowing whether the luminance value is continuously increasing or decreasing.

Next, the position of the pixel to be processed is shifted by one in step s53, and the position x of the pixel to be processed is shifted in step s54.
And the luminance value of the adjacent pixel x + 1.

At step s55, the absolute value of the difference value is equal to the threshold value T.
A pixel having a value of h1 or more is defined as a pixel whose luminance value increases or decreases. If the condition is not satisfied, it is determined that the brightness value has continuously increased or decreased.

If the condition is satisfied in step s55, the code state of the difference value is obtained in step s56, and the code state is compared with the code state stored in the flag. If the signs are the same, the process returns to step s53 to find a pixel that continuously increases or decreases. If the signs are different, it is determined that the continuous increase or decrease of the luminance value has been completed, and the pixel position x is substituted for the edge end point n2 in step s57.

In the following process, a process is performed to check how many pixels b are considered adjacent to the pixel and whose luminance is considered to be stable.

In step s58, the luminance value Y2 of the provisional edge end point is extracted.

In step s59, the pixel position is shifted by one, and in step s60, the luminance values of the pixels located at positions that are continuous with the provisional edge end point are sequentially extracted.

At step s61, it is determined whether or not the brightness value is equal to or more than Y2-β and equal to or less than Y2 + β. If the condition is satisfied, the number of pixels satisfying the condition is counted at step s62, the process returns to step s59, and the process is repeated. Find b.

On the other hand, in step s61, the luminance value is Y2-β
If not more than Y2 + β, the process ends,
The edge end point n2 and the number b of pixels for which the luminance value is considered to be stable are determined.

FIG. 8 shows the process of correcting the color difference value at the edge start point in step s10 in FIG. FIG.
In step s70, a variable c for counting the number of pixels whose color difference values continuously increase or decrease is initialized.

At step s71, the pixel position of the edge start point is set at the processing pixel position k.

At step s72, the pixel position n of the edge start point
The difference from the color difference value of the pixel shifted by 2 pixels in the direction of n1-a from 1 is calculated, and the state of the sign of the difference value is set in a flag in step s73. The flag is for determining whether the color difference value of a pixel continuously increases or decreases.

Next, at step s74, the processing pixel position is set to 1
In step s75, it is determined whether or not an area (> n1-a, <n1) is considered to be a smooth part with small luminance fluctuation. In step s75, k-2 is a region (> n1-a,
If it is determined that <n1) is deviated, the color difference value is not stabilized in the smooth portion where the luminance variation is small, and the true value of the color difference value cannot be determined, so that the color difference value is not changed.

On the other hand, if it is determined in step s75 that the pixel is within the area, in step s76, a difference between the color difference value between the position of the processing pixel and a pixel position shifted by two pixels is calculated, and step s77 is performed.
Then, the code state of the difference value of the color difference value is obtained, and the code state is compared with the code state of the flag.

If the same sign is determined in step s77, it is determined that the color difference value is continuously increasing or decreasing. In step s78, the variable c is increased by one, and the process returns to step s74. The process of counting the number of pixels for which the color difference value increases or decreases is repeated.

On the other hand, if it is determined in step s77 that the code state is different or the code state of the difference value is 0, it means that the pixels that have been continuously increasing and decreasing have ended. It is determined whether the variable c that counts the number of pixels whose value increases or decreases is 3 or more. If the variable c is 3 or more, the color difference value is at least 3
It is determined that the number of pixels has increased or decreased continuously, and the process proceeds to the next process. If the variable c is less than 3, the color difference value correction process ends. That is, in the present embodiment, the detection is performed when the continuous increase or decrease of the color difference value continues for three or more pixels.

If the variable c is 3 or more, k is set to the pixel position k1 of the end point of the area where the color difference value is changed in step s80, and the edge start point n1 from the pixel position k1 is set in step s81.
The color difference values up to are corrected.

In this embodiment, in step s81, the chrominance value of the pixel from the pixel position k1 to the edge start point is calculated as the average of the chrominance values from the pixel k1 where the continuous increase and decrease of the chrominance value ends to the end point n1-a of the luminance smoothing unit. Replace with a value. That is, the additive average value was used as the representative value C1 of the color difference values from the pixel k1 to the end point n1-a of the luminance smoothing unit.

FIG. 25 shows an example of the process of changing the color difference value from the pixel position k1 to the edge start point in step s81 in FIG.

First, in step s141, a variable c for counting the number of pixels from k1 to the edge start point n1 and a variable d for storing a value for changing the color difference value are initialized.

Next, in step s142, the pixel position k1 which is the end point of the region where the color difference value is considered to be stable is set to m.

Thereafter, the color difference value of the pixel position m is added in step s143, the pixel position m is shifted by one in step s144, and the count of the number of pixels is increased by one in step s145.

In step s146, it is determined whether or not the pixel position m of the color difference value to be obtained is within a region where the luminance is stable.
If the result of the determination is that the brightness is within the stable area, step s
Returning to 143, the color difference values are added. Thereafter, if the color difference value falls outside the stable area, an average value of the color difference values in the area where the color difference value is stable is obtained in step s147.

In step s148, the color difference values from the pixel position k1 + 1 to the edge start point n1 are replaced with the color difference values obtained in step 147.

By performing the above processing, color blur can be reduced.

Here, the edge start point n is calculated as the average of the color difference values from k1 to the end point n1-a in the luminance stable area.
Although the color difference values up to 1 have been replaced, the color difference values may be replaced with an average value of three pixels from k1 to the (n1-a) th pixel position. In this case, there is an advantage that the processing becomes faster.

FIG. 26 shows a step s141 shown in FIG.
The same operation as the process from step s147 to step s147 is performed. Then, in step s150, the color difference values up to the pixel position immediately before the edge start point are replaced with the average value d of the color difference values. In step s151, one of the edge start points is replaced. Back pixel position n1
An intermediate value between the color difference value of +1 and the average value d of the color difference values is set as the color difference value at the edge start point.

In this manner, the average value d of the color difference values is obtained.
A sharp change between the color difference value at the edge portion and the color difference value at the edge portion can be reduced, and the color blur can be smoothly reduced.

FIG. 9 shows the process of correcting the color difference value at the end point of the edge in step s11 in FIG. FIG.
In step s90, a variable c for counting the number of pixels whose color difference values continuously increase or decrease is initialized.

In step s91, the pixel position of the edge end point n2 is set to the processing pixel position k. In step s92,
2 from the pixel position n2 of the edge end point to the position of n2 + b
The difference from the color difference value of the pixel shifted by the pixel is calculated, and a step s
At 93, the state of the sign of the difference value is set in the flag. The flag is for determining whether the color difference value of a pixel continuously increases or decreases.

In step s94, the processing pixel position is shifted by one pixel, and in step s95, it is determined whether or not the area is considered to be luminance smooth (> n2, <n2 + b).
As a result of the determination, if the value is out of the area, the true value of the color difference value cannot be determined, so that the color difference value is not changed. On the other hand, in the case of the area, the difference between the color difference value of the position of the processing pixel and the pixel position shifted by two pixels is obtained in step s96, the sign state of the difference value is obtained in step s97, and the sign state is compared with the state of the flag. I do.

As a result of the comparison, if they are the same sign, it is determined that the color difference values have been continuously increasing or decreasing, and step s
At 98, the variable c for counting the number of pixels for which the color difference value continuously increases or decreases is increased by one, and the process returns to step s94 to repeat the process for counting the number of pixels for which the color difference value continuously increases or decreases. .

If the sign state of the different sign or the difference value is 0, step s99 is when the continuously increasing / decreasing pixel ends or there is no change, and the color difference value continuously increases. Alternatively, it is determined whether or not the variable c in which the number of decreasing pixels is counted is 3 or more. That is, in the present embodiment, the detection is performed by the continuous increase or decrease of the color difference value continuing for three or more pixels.

If the variable c is 3 or more, it is determined that the color difference value has increased or decreased for at least three consecutive pixels, and k is set at the pixel position k2 at the end point of the area where the color difference value is changed in step s100. The color difference values from the pixel position k2 to the edge end point are corrected in step s101.
If the variable c is less than 3, the process ends.

FIG. 27 is a flowchart showing step s101 in FIG.
3 shows an example of processing for changing the color difference value from the pixel position k2 to the edge end point.

First, in step s160, a variable c for counting the number of pixels from k2 to the edge end point n2 and a variable d for storing a value for changing the color difference value are initialized.

Next, at step s161, the pixel position k2, which is the end point of the region where the color difference value is considered to be stable, is set to m.

Thereafter, the color difference value at the pixel position m is added in step s162, the pixel position m is shifted by one in step s163, and the count of the number of pixels is increased by one in step s164.

In step s165, it is determined whether or not the pixel position m of the color difference value to be obtained is within a stable luminance region.
If the result of the determination is that the brightness is within the stable area, step s
Returning to 162, the color difference values are added. Thereafter, if the color difference value is out of the stable area, an average value of the color difference value in the area where the color difference value is stable is obtained in step s166.

In step s167, the color difference values from the pixel position k2 + 1 to the edge end point n2 are replaced with the color difference values obtained in step s166.

By performing the above processing, color blur can be reduced.

Here, the edge end point n is defined as the average of the color difference values from k2 to the end point n2 + b in the luminance stable area.
Although the color difference values up to 2 have been replaced, the color difference values may be replaced with an average value of three pixels from k2 to the n2 + b-th pixel position. In this case, there is an advantage that the processing becomes faster.

FIG. 28 is a flowchart showing step s160 shown in FIG.
The same operation as the processing from step s166 to step s166 is performed. Then, in step s170, the color difference values up to the pixel position immediately after the edge end point are replaced with the average value d of the color difference values. In step s171, one of the edge end points is replaced. Previous pixel position n2-
An intermediate value between the color difference value of 1 and the average value d of the color difference values is set as the color difference value at the edge end point.

By doing so, the average value d of the color difference values is obtained.
A sharp change between the color difference value at the edge portion and the color difference value at the edge portion can be reduced, and the color blur can be smoothly reduced.

As described above, after correcting the color difference value in the horizontal direction, the color difference value in the vertical direction is corrected next. FIG. 10 is a flowchart showing the process of correcting the color difference value in the vertical direction of the image in FIG. 4, and the detailed operation will be described below.

However, the operation described here is a case where the number of pixels in the vertical direction is processed three by three, and the processing in the vertical direction is repeated a plurality of times.

First, in step s120, a count for repeating the processing in the vertical direction a plurality of times is performed.
At 121, a position y at which processing in the vertical direction is started is set.

Next, in step s122, a page storing the luminance value of the image memory is set, and in step s123, the luminance values at the pixel positions y-1, y, y + 1 are sequentially extracted from the image memory in three vertical pixels.

In step s124, the luminance value Yh at the pixel position y and the upper and lower pixel positions y-1, y adjacent to the pixel position y
A pixel having a luminance value of a pixel of +1 having a luminance value of Yh-δ or more and Yh + δ or less is selected from upper and lower pixels.

In step s125, the page in which the color difference values of the image memory 8 are stored and the page for temporarily storing the color difference values in the image line memory 10 (see FIG. 2) are switched. In step s126, the color difference value of the pixel position y is changed. The average value of the pixel value and the color difference value of the pixel selected in step s124 is output to the image line memory 10 as the color difference value at the pixel position y.
At this time, if there is no pixel selected in step s124, the color difference value is not changed. Therefore, the color difference value at the pixel position y is stored in the image line memory 10 as it is.

Next, in step s127, in order to perform processing for the other color difference value, it is determined whether to switch pages of the image memory 8 and the image line memory 10.

If the processing has not been completed for both color difference values, in step s128, the pages of the image memory 8 and the image line memory 10 are set to 2, and the above processing is repeated to terminate the processing for both color difference values. Then, in step s129, it is determined whether the processing in the vertical direction has been completed.

If the processing has not been completed, the processing pixel position y is increased by one in step s130, the processing is returned to step s122, and the above processing is repeated.

On the other hand, if it is determined in step s129 that the processing in the vertical direction has been completed, the corrected color difference value stored in the image line memory 10 is transferred to the corresponding pixel position in the image memory 8 in step s131. .

Next, in step s132, it is determined whether the processing in the vertical direction has been repeated a preset number of times, and the processing is repeated for the desired number of times. In step s133, when the processing in the vertical direction has been completed for the set number of times, it is determined whether the processing has been performed on all the images. If the processing has not been completed, the processing pixel position in the horizontal direction is increased by one in step s134, the process returns to step s120, and the processing is repeated to perform processing on all images to reduce color noise on all images. .

In this embodiment, the color difference values from the edge end point to the k2th pixel are replaced with the average of the color difference values from the k2th pixel to the k2 + bth pixel. That is, in the present embodiment, the additive average value is used as the representative value C2 of the color difference values from the k2th pixel to the k2 + bth pixel.

In addition, 256 gradations were set, with the lowest luminance value being 0 level and the highest luminance value being 255 levels. Then, it was found that it is preferable to set the threshold value Th1 between 3 and 10. This is because the purpose is to find a pixel whose brightness value continuously increases or decreases, but it is difficult to obtain a brightness value that continuously increases or decreases because the image contains noise. Therefore, in the present invention, in consideration of image noise, a difference value of a luminance value is set to a pixel in which the difference value of a pixel outside a certain range continuously increases or decreases.

That is, if the threshold value Th1 is 2 or less, it is difficult to discriminate an area which is considered to continuously increase or decrease due to noise. Therefore, the effect of reducing color blur is reduced. From the above,
It is considered that it is preferable to set the value of the threshold Th1 to a value of about 1% to 4% of the maximum luminance value.

It was also found that it is preferable to set α and β between 3 and 10. This is because the purpose is to find continuous pixels whose luminance value is considered to be constant, but it is difficult to obtain a constant luminance value because the image contains noise. Therefore, in the present invention, in consideration of image noise,
Pixels whose luminance values are within a predetermined constant range are defined as pixels whose luminance values are constant. That is, if α and β are 2 or less, it is difficult to determine a region considered to be stable due to noise. If α and β are 11 or more, an extra region in which the luminance value changes is often determined to be a region considered to be stable. This is because of the From the above, it is considered that it is preferable to set the values of α and β to a value of about 1% to 4% of the maximum luminance value.

The values of α and β can be changed in proportion to the magnitudes of the luminance values Y1 and Y2. For example, a value of Z% (<50%) of the luminance value, or several kinds of fixed values can be allocated in proportion to the magnitude of the luminance value.

Next, in the correction of the color difference value, in this embodiment, the difference between the position of the processing pixel and the color difference value of the pixel position shifted by two pixels is obtained. This is because the change in the color difference signal is gentle, and when the difference value between adjacent pixels is calculated, the change in the color difference value is difficult to understand. In order to make it easy to understand the continuously increasing or decreasing value, a color difference value at a pixel position shifted by two pixels is taken. Of course, the difference value may be obtained by shifting by three or more pixels.

In this embodiment, the vertical chrominance value correcting means 6 performs the processing on the pixels consecutive in the vertical direction a plurality of times.

In this embodiment, the number of pixels in the vertical direction is three.
Although the processing is performed pixel by pixel, it is needless to say that the processing can be performed while reading out the odd number of pixels of 5 pixels or more. Further, it is difficult to refer to the same number of pixels above and below the pixel of interest at the periphery of one image. Therefore, in this area, refer to pixels that are not symmetrical above and below (there may be an even number of pixels). Can also.

In this embodiment, 256 gradations are set, with the lowest luminance value being 0 level and the highest luminance value being 255 levels. Then, it was found that it is preferable to set δ between 3 and 10. This is because the purpose is to find a smooth portion of the luminance value. However, if the value is set to be smaller than 3, there is often no pixel to be selected for noise, and the color noise can be efficiently smoothed. become unable.

If the value is larger than 10, even if the color changes at the luminance edge portion, the averaging process is performed on the different color portion without finding the edge portion, so that the color noise cannot be reduced. From the above, the value of δ is set to 1 of the maximum brightness value.
It is desirable to set the value to about 4%. The value of δ used in this embodiment is set to 8.

The number of repetitions in the vertical direction is preferably set from 3 to 10. This is because the number of times of processing is to increase the number of pixels to be referred to as a whole and to further reduce color noise. However, if it is performed twice or less, the number of reference pixels is small, and it is difficult to reduce color noise. If it is performed ten or more times, the number of reference pixels increases and the color becomes lighter as a whole. In this embodiment, the number of repetitions is set to eight.

By performing the above-described color difference value correction on the entire image data by the vertical color difference value correction means 6, color noise slightly generated by the color bleeding process can be reduced, and a processed image of good image quality can be obtained. it can.

In this embodiment, the processed chrominance values are stored in the image line memory 10, and the pixels in the vertical direction are processed by one line and then transferred to the image memory 8 at once. Then, the image data may be transferred to the image memory 8 when the processing of the pixel position to be processed is completed. Here, the image line memory 10 is used for easy understanding of the processing.

By the above color difference value processing, color blur is considerably suppressed as compared with the original image, and the image quality can be sufficiently improved. However, when the processing of the edge start point color difference value correction means and the edge end point color difference value correction means is performed by the above-described correction processing, the edge suddenly changes more than the gentle change of the color difference signal.
The image is partially emphasized at the contour portion where the color of the image changes, and the image tends to be slightly unnatural.

Therefore, as a second embodiment, the color difference values up to the edge start point and the edge end point replaced by the average of the color difference values in the first embodiment are further compared with the color difference value of the replaced edge start point and the edge end point. Are replaced by linear interpolation.

By doing so, the edge portion and the edge of the color difference value coincide with each other, and color fringing is removed, partial edge enhancement of color does not occur, and good image quality without giving an unnatural impression is obtained. To obtain a good recorded image.

Table 1 shows numerical data of correction results according to the first embodiment and the second embodiment immediately after performing horizontal processing on the luminance signal and the color difference signal of the original image. . In this experimental example, color difference value correction processing was performed at Th1 = α = β = 5. Each row indicates a pixel position, column A indicates a pixel position number, column B indicates the luminance value of the original image, column C indicates the color difference value of the original image, and column D indicates the color difference value of the image processed in the first embodiment. Column E shows the color difference values of the image processed in the second embodiment,
The columns indicate each pixel detected during processing. However,
Although there are two signals of RY and BY signals in the color difference value,
Since the processing is the same, only the BY signal is shown.

[0152]

[Table 1]

FIG. 11 is a graph showing the result of the processing in (Table 1). The horizontal axis indicates the pixel position, and the vertical axis indicates the luminance value or color difference value of each pixel. FIG. 11A shows a luminance value of an original image, where n1 is a luminance edge start point,
n2 indicates an edge end point, and n1-a and n2 + b indicate end points of a luminance smoothing portion. FIG. 11B shows color difference values corresponding to pixel positions of luminance values of the original image. k
1 and k2 indicate the end points of the color fringing portion based on the color difference value. FIG. 11C shows a result obtained by performing the process according to the first embodiment, and FIG. 11D shows a result obtained by the second embodiment. In the result shown in FIG. This is the result of changing the color difference value of.

In the processing result shown in FIG. 11C, it can be confirmed that the rounding of the color difference value has been improved and the color blur has been removed. However, it can be seen that the color difference values change discontinuously at the edge start point and the end point, and the color change is slightly emphasized.

In FIG. 11D, the color difference values between the edges are connected by linear interpolation to clarify the change in the color difference values, thereby clarifying the edges with respect to the color difference values and making the color difference values discontinuous. It does not change. Therefore, the color blur is removed in the same manner as in FIG.
Further, it can be seen that an image in which the color change is not unnecessarily emphasized is obtained.

FIG. 12 is a graph showing experimental results obtained by correcting color difference values of certain image data immediately after horizontal color difference value correction processing according to the second embodiment and thereafter by performing vertical color difference value correction means. The vertical axis indicates the luminance value or the color difference value of each pixel value, the horizontal axis indicates the pixel position in the horizontal direction, and the diagonal axis direction indicates the pixel position in the vertical direction.

FIG. 12A shows the luminance value of the original image. The point where the luminance becomes a uniform value corresponds to the end point of the luminance.

FIG. 12B shows the color difference values at the pixel positions corresponding to FIG. The color difference value changes gradually in a portion where the luminance value spreads uniformly. This gentle portion is a portion where color blur has occurred.

FIG. 12C shows the result immediately after the correction processing of the color difference value in the horizontal direction in order to remove the color fringing portion. In the middle part, a part that could not be processed as color blur appears. This is because the region where the color difference value changes continuously is larger than the region where the luminance value is smooth.

FIG. 12D shows the result of performing the correction process of the color difference value in the vertical direction. In FIG. 12 (c), the portion where the color blur cannot be removed is removed. This is because the peripheral pixels have a uniform distribution of the luminance values of the pixels in the vertical direction, and it is assumed that the chrominance signals are also the same, and are not removed by averaging the chrominance values. This is because the blurred color portion is distributed to surrounding pixels and smoothed. As a result, the color blur is removed and the image quality is improved, and the variation in the color difference value in the vertical direction is also improved, thereby improving the image quality.

Here, the pixels n1 to n2 are the luminance change areas, and the pixels n1 to n1-a and the pixel n2
To n2 + b is a luminance stable area, and pixels n1 to k1 and pixels n2 to k2 are color difference value change areas.

As described above, first, the edge start point detecting means 2 and the edge end point detecting means 3 detect an edge having a luminance value whose luminance continuously changes in the horizontal scanning direction. next,
With respect to the chrominance value changing around the edge of the detected luminance value, the chrominance value correcting means 4 at the edge start point and the chrominance value correcting means 5 at the edge end point change the chrominance value in accordance with the change in the luminance value. Change the color difference value to Further, the vertical color difference value correcting means 6 extracts pixels that are considered to be continuous in the vertical direction of the image and corrects the color difference values. That is, in the horizontal direction and the vertical direction of the image, the color difference signal is corrected in two stages so that the change in the color difference value is matched with the change in the luminance value, so that the rounding of the color difference signal is corrected. Color bleeding can be greatly reduced.

In a general image, a video signal is transmitted in the horizontal scanning direction. In particular, the color difference signal becomes large due to the band limitation of the video signal, and the color difference value cannot follow the change in the luminance value, and is observed as a color blur. Most of it. In other words, the part where the color changes is correlated with the part where the luminance value changes,
In particular, where the change in the luminance value is large, the change in the color difference signal is also large, and is observed as a large color blur.

Accordingly, first, a portion where the color value is most noticeable in the horizontal scanning direction and a change in the luminance value is large (that is, an edge portion) is detected, and the color blur at this portion is removed as much as possible.
However, when the correction is performed only in the horizontal direction, the portion from which color blur is removed and the portion that cannot be removed may be vertically adjacent to each other, and conversely, an unnatural image may be generated and the image quality may be degraded. . Therefore, if the chrominance values are further corrected in the vertical direction using the luminance values, the chrominance values in the vertically adjacent portions are averaged, so that the chrominance values are prevented from changing discontinuously and uniform. Thus, the color blur can be reduced, and the image quality can be greatly improved with a natural image.

Next, a video signal correction device according to the third embodiment shown in FIG. 13 will be described.

In FIG. 13, the components having the same numbers as those in FIG. 1 are the same. Reference numeral 11 denotes an access direction switch for switching the reading direction of the luminance value and the color difference value from the image memory 1 and the storing direction thereof to either the horizontal forward direction or the horizontal reverse direction of the image.

In the above configuration, first, the luminance value of the image is extracted from the image memory 1 in the access direction determined by the access direction switch 11 in either the horizontal forward direction or the horizontal reverse direction, and the edge start point is detected. The means 2 detects an edge start point which is the n1th pixel at which the luminance value starts to largely change, and the edge end point detecting means 3 detects an edge end point which is the n2th pixel at which the change of the luminance value ends.

Next, while the color difference values of the image are taken out from the image memory 1 in the access direction and the color difference values are continuously changing immediately after the edge end point, the color difference value is corrected by the edge end point color difference value correcting means 5. The change is performed for all the image data.

Next, the access direction switch 11
The same processing is performed in a direction opposite to the access direction processed previously using the edge start point detection means 2, the edge end point detection means 3, and the edge end point color difference value correction means 5, and the color difference in the vicinity where the luminance value greatly changes is obtained. Correct the value.

Next, the luminance value of the image is taken out from the image memory 1 in the vertical direction, and the chrominance value of the pixel whose luminance is smooth is averaged by the vertical chrominance value correcting means 6 as described above. Forward. This averaging process is repeated a plurality of times.

With the above-described configuration, it is possible to obtain an image with improved image quality in which not only color fringing is removed but also color noise is reduced, as in the first embodiment. Particularly, in this configuration, it is only necessary to change the processing direction, and the processing configuration is smaller than that of the embodiment shown in FIG.

In the present embodiment, in the correction of the color difference value, the entire image is processed in either the horizontal forward direction or the horizontal reverse direction, and then the access direction opposite to the processed access direction is performed. By processing in the access direction, a processed image of color difference value correction is obtained. However,
Similar results can be obtained by alternately performing horizontal forward processing and horizontal reverse processing for each line of the image.

The configuration of the image data processing apparatus in which this embodiment is realized by a microcomputer is the same as that of FIG. However, in the present embodiment, the microcomputer 8 performs the edge start point detection means 2, the edge end point detection means 3, the edge end point color difference value correction means 5, the access direction switch 11 and the vertical color difference Each processing of the operation of the value correction means 6 is performed.

In the video signal processing apparatus configured as described above, the operations of the edge start point detecting means 2, the edge end point detecting means 3, the edge end point color difference value correcting means 5, and the vertical color difference value correcting means 6 are the same as those described above. Since the operation is the same as that of the first embodiment, details are omitted in the following description of the operation.

FIG. 14 shows the relationship between the position of the image memory and the position of the pixel used in this embodiment. The horizontal scanning direction of the video signal coincides with the horizontal forward direction (x direction) of the image memory, the direction opposite to the horizontal forward direction is the horizontal reverse direction, and the sub-scanning direction coincides with the vertical direction (y direction) of the image memory. I have. (X, y) represents a pixel position in the image, and corresponds to each address of the image memory.

FIG. 15 is a flowchart showing horizontal processing in the video signal correction device shown in FIG. 13, and each processing will be described below. Note that the components having the same numbers as those in FIG. 5 are the same. However, the horizontal direction described below follows the direction set by the access direction switch 11.

First, in step s150, the access direction of the image memory is set, and the access direction switch 11
Has been realized. In step s3, the position of a pixel in the image memory 8 at which processing in the vertical direction is started is set. In step s4, the position of a pixel in the image memory 8 at which horizontal processing is started is set. In step s5, a page for extracting a luminance value from the image memory 8 is set. In step s6, the luminance values are sequentially extracted from the image memory 8, and
The pixel position n1 of the edge start point of the luminance value is obtained. This processing implements the edge start point detecting means 2.

In step s7, the detected edge start point 3
It is checked whether the pixel ahead of the pixel does not exceed the horizontal end point of the image. When the horizontal end point is exceeded, the edge end point cannot be detected, so that the color difference value is not corrected.

In step s8, the luminance value is extracted from the image memory 8, the edge end point of the luminance value is found, and the position information n2 and the luminance value of the pixel after the edge end point continuously increase or decrease by at least two pixels or more. The pixel position n2 + b to be started is obtained, and the process from the pixel n2 to the n2 + b-th pixel is determined as an area having a small variation in luminance value. This processing implements the luminance edge end point detecting means 3.

In step s9, the page of the image memory 8 is switched to read the color difference value, and the color difference value R−
A setting for extracting a color difference value of either Y or BY is performed. In step s11, the chrominance values of the pixels in the region where the fluctuation of the luminance value obtained in step s8 is small are stored in the image memory 8
Are sequentially taken out from the storage position, the k2th pixel is detected, the color difference values from the edge end point to the k2th pixel are corrected, and output to the image memory 8. This processing implements the edge end point color difference value correcting means 5.

In step s12, the respective color difference values RY, B-
A determination is made to switch the page of the image memory 8 for performing color difference value correction for Y. In step s13, it is determined whether the pixel position to be processed has been processed for one line in the horizontal direction of the image. If there are unprocessed parts,
In order to correct the color difference values for all the pixels in the horizontal direction, the process returns to step s5 and the process is repeated.

At step s14, it is determined whether or not the video signal correction processing has been completed for the entire image to be processed. If the processing has not been completed, the address y in the vertical direction is incremented by one in step s15, and the process returns to step s4 to repeat the processing in the horizontal direction.

In step s151, it is determined whether or not the processing in the horizontal forward direction and the horizontal reverse direction has been completed for the entire image to be processed. If the process has not been completed, the process returns to step s150, and the above process is repeated by reversing the access direction of the pixel value to be extracted.

As described above, by performing the color difference value processing in the horizontal direction, and then executing the vertical color difference value correction means, the color blur can be suppressed as compared with the original image, and the image quality can be improved. However, in the processing in the horizontal direction, the edge changes abruptly more than the gentle change of the color difference signal. Therefore, depending on the location, the image is partially emphasized at the outline where the color changes, and the image tends to be slightly unnatural.

Therefore, in the above-described color difference value correction, the color difference value at the edge end point n2 is changed to the average value of the (n2-1) th pixel value and the representative value C2 of the color difference values in a region where the color difference value is less varied. As a result, it is possible to reduce partial color edge enhancement at the edge portion and the edge portion of the color difference value, and it is possible to obtain a better recorded image that suppresses an unnatural impression. This is the fourth embodiment.

Table 2 shows numerical data of experimental results immediately after the luminance signal and the color difference signal of a certain original image are processed in the third and fourth embodiments in the horizontal direction. .
In the third and fourth embodiments, the edge end point detecting means 3
In order to detect the pixel at the end point n1 + b-th in the area where the luminance value does not fluctuate continuously to the edge end point, the pixel of the edge start point that appears immediately after the edge end point currently being searched is set to n2.
The pixel is determined as the + b-th pixel.

In this experimental example, the threshold value used in the edge start point detection means and the edge end point detection means was set to Th1 = 5 for processing. Column A shows the pixel position, column B shows the luminance value of the original image, column C shows the color difference values of the original image, column D shows the color difference values processed in the horizontal forward direction, and column E shows the values after the processing in the horizontal forward direction. It shows the color difference values when processing is performed in the horizontal reverse direction. Row F is the fifth
Shows the color difference values processed in the horizontal forward direction of the image processed in the embodiment, and the G column shows the color difference values obtained when the image is processed in the horizontal reverse direction after the processing in the horizontal forward direction. However, the color difference value is R-
Although there are two signals, Y and BY signals, only the BY signal is shown because the processing is the same.

[0188]

[Table 2]

FIG. 16 is a graph showing the experimental results in (Table 2). The horizontal axis indicates the pixel position, and the vertical axis indicates the luminance value or color difference value of each pixel. 16A shows the luminance value of the original image, FIG. 16B shows the color difference values corresponding to the pixel positions of the luminance value of the original image, and FIG. 16C shows the third embodiment. FIG. 1 shows the result of processing in the horizontal forward direction in the example.
6 (d) shows the result of processing in the horizontal reverse direction after processing in the horizontal forward direction in the third embodiment.

In FIG. 16C, the rounding of the color difference values on only one side in the horizontal forward direction is improved, and in FIG. 16D, the rounding of the color difference values is improved as in the first embodiment. That is, it can be seen that by correcting the color difference value in which the color difference value continuously increases or decreases from the edge, the color blur is removed, and the image quality is greatly improved. However, the color difference value changes discontinuously at the edge, and the color change is slightly emphasized at the edge.

FIGS. 16E and 16F are graphs showing the processing results according to the fourth embodiment described above. The color difference values do not change discontinuously by interpolating the discontinuous color difference values near the edge using the average value.
Therefore, the color bleeding is the same as the processing in the second embodiment.
It can be seen that a corrected image with a natural image quality is obtained, which is removed in the same manner as in (d), and further, the color change is not unnecessarily emphasized.

Further, by performing the processing by the vertical color difference value correcting means, it is possible to satisfactorily remove the color fringing even in the part incompletely processed in the horizontal direction as in the first embodiment.

FIG. 17 is a block diagram showing the configuration of a video signal correction device according to the fifth embodiment of the present invention. The components having the same numbers as those in FIGS. 1 and 13 are the same as those described above.

In FIG. 17, first, from the image memory 1 in the access direction determined by the access direction switch 11 in one of the horizontal forward direction and the horizontal reverse direction,
The luminance value of the image is taken out, and the edge start point detecting means 2 detects the edge start point which is the n1st pixel at which the luminance value starts to change greatly.

Next, while the color difference values of the image are fetched from the image memory 1 in the access direction and if the color difference values are continuously changing immediately before the edge start point, the edge start point color difference value correction means 4 causes the area to change. Change the color difference value at
By repeating this operation, the color difference value correction for one image data in the horizontal direction is completed.

Next, the access direction switch 11
This time, the same processing is performed in the opposite direction to the processed access direction, and the color difference value in the vicinity where the luminance value greatly changes is corrected.

Next, the luminance value of the image is taken out from the image memory 1 in the vertical direction, and the vertical chrominance value correcting means 6 averages the chrominance values of the pixels whose luminance is smooth and averages them.
Transfer to The above processing is performed for all the image data to obtain a processed image with reduced color noise.

The present embodiment is the simplest of the above-described examples of the configuration of the video signal correction device, and can be realized at the lowest cost.

In the correction of the color difference value, the entire image is processed in either the horizontal forward direction or the horizontal reverse direction, and then the processing in the access direction opposite to the processed access direction is performed. , A processed image is obtained, but the same result can be obtained by alternately performing the processing in the horizontal forward direction and the processing in the horizontal reverse direction for each line of the image.

The configuration of an image data processing apparatus in which this embodiment is realized by a microcomputer is the same as that shown in FIG. However, the microcomputer 9 performs the respective processes of the operations of the edge start point detection means 2, the edge start point color difference value correction means 4, the access direction switch 11, and the vertical color difference value correction means 6 on the input video signal. be able to.

In the video signal processing apparatus configured as described above, the operation of the edge start point detection means 2, the edge start point color difference value correction means 4, and the vertical color difference value correction means 6 is the same as that of the first embodiment and the second embodiment. The description is omitted because it is the same as that of the embodiment, and the operation will be described below. Here, the relationship between the position of the image memory 8 and the position of the pixel is the same as in FIG.

FIG. 18 is a flowchart showing processing in the horizontal direction of the video signal correction device. Each processing will be described below. The components having the same numbers as those in FIGS. 5 and 15 are the same. However, the horizontal direction described below follows the direction set by the access direction switch 11.

First, in step s150, the access direction of the image memory is set, and the access direction switch 11 is realized. In step s3, the position of a pixel in the image memory 8 at which processing in the vertical direction is started is set.
In step 4, the position of a pixel in the image memory 8 at which horizontal processing is started is set. In step s5, a page for extracting a luminance value from the image memory 8 is set.

In step s6, the luminance values are sequentially taken out from the image memory 8, and the pixel position n1 of the edge start point of the luminance value is obtained. This processing implements the edge start point detecting means 2. A step s7 checks whether or not the pixel three pixels ahead of the detected edge start point does not exceed the horizontal end point of the image. When the horizontal end point is exceeded, it is not possible to determine whether the change in luminance starting from the edge start point is an edge that continuously increases or decreases, so that the color difference value is not corrected.

In step s9, the page of the image memory 8 is switched to read the value of the color difference value, and the color difference value R is read.
A setting for extracting a color difference value of either -Y or BY is performed.

In step s10, the color difference values of the pixels in the area where the luminance value does not vary much, obtained in step s6, are sequentially extracted from the storage position in the image memory, and the k1st pixel is detected. The color difference values up to the start point are corrected and output to the image memory 8. This processing implements the edge start point color difference value correcting means 4.

In step s12, the respective color difference values RY, B-
A determination is made to switch the page of the image memory for performing color difference correction for Y. In step s13, it is determined whether the pixel position to be processed has been processed for one line in the horizontal direction of the image. If there is an unprocessed portion, the process returns to step s5 and repeats the process to correct the color difference value for all pixels in the horizontal direction.

At step s14, when the horizontal processing is completed at step s13, it is determined whether or not the video signal correction processing has been completed for the entire image to be processed. If the processing has not been completed, the address y in the vertical direction is incremented by one in step s15, and the process returns to step s4 to repeat the processing in the horizontal direction.

In step s151, when the processing in the vertical direction is completed in step s14, it is determined whether or not the processing in both the horizontal forward direction and the horizontal reverse direction is completed for the entire image to be processed. If the process has not been completed, the process returns to step s150, the access direction of the pixel value to be taken out is set, and the process is repeated.

As described above, by performing the vertical color difference value correction means after the horizontal color difference value processing, color blur and color noise are considerably suppressed as compared with the original image, and the image quality can be improved.

In this embodiment, the edge start point color difference value correcting means 4 converts the color difference value from the edge start point detected by the edge start point detecting means 2 to the (n1-a) th pixel of the area where the luminance value does not change much. Take out sequentially from the storage location of the memory,
The k1th pixel at which the color difference value continuously increases or decreases is detected, and the median color difference value from the k1st pixel to the n1-ath pixel is calculated as the color difference value from the k1st pixel to the edge start point. Is output to the image memory 8.

Next, as a sixth embodiment, similar to the fifth embodiment, the color difference value of the edge start point n1 is changed to the representative value C1 of the color difference value in an area where the change of the color difference value is small with the (n + 1) th pixel value.
, It is possible to reduce partial color edge enhancement at the edge portion and the edge portion of the color difference value, and it is possible to obtain a good recorded image that suppresses an unnatural impression.

Next, by executing the vertical color difference value correcting means, portions which are not sufficient only by the horizontal color blur correction processing are improved, and a better recorded image can be obtained.

Table 3 shows processing result data according to the fifth and sixth embodiments for the luminance signal and the color difference signal of the original image. In this experimental example, the threshold value used in the edge start point detection means 2 was set to Th1 = 5, and the processing was performed.
A indicates the pixel position, B indicates the luminance value of the original image, C indicates the color difference value of the original image, D indicates the color difference value processed in the horizontal forward direction of the fifth embodiment, and E indicates the value of the horizontal forward direction processing. Thereafter, the color difference values when the processing is performed in the horizontal reverse direction are shown.

F indicates the color difference value of the image processed according to the sixth embodiment in the horizontal forward direction, and G indicates the color difference value of the image processed in the horizontal reverse direction after the horizontal forward process. ing. However, there are two signals of the RY and BY signals in the color difference value, but since the processing is the same, only the BY signal is shown.

[0216]

[Table 3]

FIG. 19 is a graph showing the processing result in (Table 3).

In FIG. 19, the horizontal axis represents a pixel position, and the vertical axis represents a luminance value or a color difference value of each pixel. FIG. 19 (a)
Represents the luminance value of the original image, and FIG. 19B shows the color difference values corresponding to the pixel positions of the luminance value of the original image. FIG.
FIG. 9C shows the result of horizontal forward processing in the fifth embodiment, and FIG. 19D shows horizontal reverse processing after horizontal forward processing in the fifth embodiment. The results obtained are shown.

In FIG. 19C, the dullness of the color difference values on only one side in the horizontal forward direction is improved, and in FIG. 19D, as in the first embodiment, the dullness of the color difference values is improved. That is, it can be seen that by correcting the color difference value in which the color difference value continuously increases or decreases from the edge, the color blur is removed, and the image quality is greatly improved.

FIGS. 19E and 19F are characteristic diagrams of processing results according to the sixth embodiment. As in the fourth embodiment,
It can be seen that a corrected image having a natural image quality in which a change in color is not unnaturally emphasized is obtained.

Next, the configuration of a video signal correction device according to the seventh embodiment of the present invention is the same as that of FIG.

Here, 6 reads out the luminance value and the color difference value of at least three pixels continuous in the vertical direction from the image memory 1 in which the color difference value corrected in the horizontal direction is stored,
One of the pixels is regarded as a target pixel, and the absolute value of the difference between the luminance values of the target pixel is the Y value of the luminance value Yh.
% (Y is a value of 5 or more and 30 or less) or less is selected from the remaining pixels, and the color difference value of the target pixel is changed using the color difference value of the selected pixel.

Next, the operation of the above embodiment will be described.
However, the processing in the horizontal direction is the same as that in the first embodiment, and the description is omitted.

Now, a luminance value of a total of three pixels is taken out from the image memory 1 storing the image for which the horizontal processing has been completed, and one pixel above and below the pixel of interest, and the vertical color difference value correction is performed. Means 6 is that the difference from the luminance value of the remaining pixel is Y% based on the luminance value Yh of the target pixel.
A pixel that satisfies (5 ≦ Y ≦ 30) or less is selected, and the color difference value of the pixel of interest is corrected using the color difference value of the selected pixel and output. In this way, the color difference values of the pixels that are continuous in the vertical direction are corrected, and the color difference values are transferred to the image memory 1 when the correction is completed.

The above processing is performed for all the image data, and a processed image with reduced color fringing and color noise can be obtained.

FIG. 20 is a flowchart showing the flow of processing of the vertical color difference value correcting means 6 in the seventh embodiment, and the detailed operation will be described below.

However, the operation described here is a case where the number of pixels in the vertical direction is processed three by three and the processing in the vertical direction is repeated a plurality of times. Further, when the step numbers are the same as those in FIG. 10, the processing is the same.

First, in step s120, counting for repeating the processing in the vertical direction a plurality of times is performed.
At 121, a position y at which processing in the vertical direction is started is set.

Next, in step s122, a page in which the luminance value of the image memory is stored is set. In step s123, the luminance values at the pixel positions y-1, y, y + 1 are sequentially extracted from the image memory by three pixels in the vertical direction.

At step s140, the luminance value Yh at the pixel position y and the upper and lower pixel positions y−1,
A pixel whose absolute value of the luminance difference of the pixel of y + 1 is a value within Y% of Yh is selected from upper and lower pixels.

In step s125, the page storing the color difference values in the image memory 8 and the image line memory 10
The page for temporarily storing the color difference value (see FIG. 2) is switched, and the average value of the color difference value of the pixel position y and the color difference value of the pixel selected in step s124 is set as the color difference value of the pixel position y in step s126. , To the image line memory 10. At this time, if there is no pixel selected in step s140, the color difference value is not changed. Therefore, the color difference value at the pixel position y is stored in the image line memory 10 as it is.

Next, in step s127, in order to process the other color difference value, it is determined whether the pages of the image memory 8 and the image line memory 10 are switched.

If the processing has not been completed for both color difference values, in step s128, the pages of the image memory 8 and the image line memory 10 are set to 2, and the above processing is repeated to terminate the processing for both color difference values. Then, in step s129, it is determined whether the processing in the vertical direction has been completed.

If not over, step s130
Then, the processing pixel position y is increased by one, the processing is returned to step s122, and the above processing is repeated.

On the other hand, if it is determined in step s129 that the processing in the vertical direction has been completed, the corrected color difference value stored in the image line memory 10 is transferred to the corresponding pixel position in the image memory 8 in step s131. .

Next, in step s132, it is determined whether the processing in the vertical direction has been repeated a predetermined number of times, and the processing is repeated for the desired number of times. In step s133, when the processing in the vertical direction has been completed for the set number of times, it is determined whether the processing has been performed on all the images. If the processing has not been completed, the processing pixel position in the horizontal direction is increased by one in step s134, the process returns to step s120, and the processing is repeated to perform processing on all images to reduce color noise on all images. .

In the setting of Y% in the vertical color difference value correcting means 6, if the luminance value of the pixel of interest is small, a change in color is difficult to be recognized due to human visual effects. It is necessary to widen the difference range from the pixel.

Therefore, the absolute value of the difference between the luminance value and the luminance value Yh of the target pixel is set to a value of Y% (5% to 30%) of Yh.

When Y% is 5% or less, almost no upper and lower pixels are selected due to noise, and the processing effect is weak. Also, where the difference in luminance value is large, the change in color difference value is also large in many cases, so if Y% is larger than 30%,
This is because there is a risk of causing color blur in the vertical direction.

In this embodiment, Y% is set to 10%, processing is performed using a total of three pixels using upper and lower pixels adjacent to the target pixel, and the average value of the color difference value of the selected pixel and the color difference value of the target pixel is obtained. Is output as a new color difference value of the pixel of interest.

In this embodiment, the vertical chrominance value correcting means 6 corrects the chrominance value by referring to three consecutive pixels. However, the vertical chrominance value correcting means 6 refers to an odd number of three or more pixels, for example, five pixels and seven pixels. It can also be done. However, in this case, the pixels are selected so that at least the luminance value is not less than Y% of the luminance value of the pixel of interest and a pixel group that is continuous in position.

Although the above-described processing is performed a plurality of times on pixels that are continuous in the vertical direction, the processing may be performed once.

The number of repetitions in the vertical direction is preferably set from 3 to 10. This is because the number of times of processing is to increase the number of pixels to be referred to as a whole and to further reduce color noise. However, if the number is less than two, the number of reference pixels is small, and it is difficult to reduce color noise. Therefore, in this embodiment, the number of repetitions is set to four.

By performing the above-described color difference value correction on the entire image data by the vertical color difference value correction means 6, color noise slightly generated due to color blurring processing can be reduced, and a processed image of good image quality can be obtained. it can.

In this embodiment, the processed chrominance values are stored in the image line memory 10, and the pixels in the vertical direction are processed by one line and then transferred to the image memory 8 at once. Then, the image data may be transferred to the image memory 8 when the processing of the pixel position to be processed is completed. Here, the image line memory 10 is used for easy understanding of the processing.

FIG. 21 is a graph showing experimental results obtained by processing certain image data according to the present embodiment. The ordinate indicates the luminance value or the color difference value of each pixel value, and the abscissa indicates the horizontal value. Direction, and the diagonal axis direction indicates the pixel position in the vertical direction.

FIG. 21A shows the luminance value of the original image. The point where the luminance becomes a uniform value corresponds to the end point of the luminance.

FIG. 21B shows color difference values at pixel positions corresponding to FIG. 21A. The color difference value changes gradually in a portion where the luminance value spreads uniformly. This gentle portion is a portion where color blur has occurred.

FIG. 21 (c) shows the result after the color difference value correction processing in the horizontal direction in order to remove the color fringing portion. In the middle part, a part that could not be processed as color blur appears. This is because the region where the color difference value changes continuously becomes larger than the region where the luminance value is smooth.

FIG. 21D shows the result of performing the correction process of the color difference value in the vertical direction on the image of FIG. 21C. In FIG. 21 (c), the portion where the color blur cannot be removed is removed. This is because the peripheral pixels have a uniform distribution of the luminance values of the pixels in the vertical direction, and it is assumed that the chrominance signals are also the same, and are not removed by averaging the chrominance values. This is because the blurred color portion is distributed to surrounding pixels and smoothed. As a result, the color blur is removed and the image quality is improved, and the variation in the color difference value in the vertical direction is also improved, thereby improving the image quality.

Next, an image signal processing apparatus according to an eighth embodiment of the present invention will be described with reference to the drawings.

FIG. 22 is a block diagram showing the configuration of a video signal correction apparatus according to the eighth embodiment of the present invention, in which the same numbers are the same as in FIG. 12
Is a low-pass filter processing unit that performs low-pass filter processing on the luminance value read from the image memory 1.

In the above configuration, video signal processing is performed as follows.

First, the luminance value of the image is taken out from the image memory 1 in the horizontal direction, and the luminance value whose noise has been reduced by the low-pass filter processing means 12 is taken out. The edge start point, which is the pixel of, and the edge end point, which is the n2th pixel at which the change of the luminance value ends, are detected by the edge end point detecting means 3.

Next, when the color difference value of the image is taken out from the image memory 1 in the horizontal direction and the color difference value is continuously changing immediately before the edge start point or immediately after the edge end point,
The edge start point color difference value correction means 4 and the edge end point color difference value correction means 5 change the color difference value in each area. By this operation, the blur portion can be removed as much as possible while detecting the color blur in the horizontal direction of the image.

Further, the color difference value corrected so as to reduce the color blur in the horizontal direction is corrected by the vertical color difference value correcting means 6 in the vertical direction, thereby further reducing the color blur. A processed image can be obtained.

That is, the color blur can be reduced in the same manner as described in the first embodiment. However, in this embodiment, since the low-pass filter processing means 12 is added, the noise included in the luminance can be reduced in advance. Thereby, the edge detection and the detection of the luminance smooth region are facilitated.

In this embodiment, the low-pass filter means 12 will be described using a median filter. However, the same effect can be obtained by using a general smoothing filter such as an averaging filter using an average value of several pixels. Can be obtained.

Since the low-pass filter processing means 12 must refer to the color difference value before being processed when being referred to as the processing for the adjacent pixel, the luminance value of the already processed pixel is stored in the image memory 1. It also has a function of temporarily storing data so that it is not written to the memory.
This can be composed of, for example, a line memory for temporarily storing the processing results in the horizontal direction, and a plurality of buffers having a capacity that can be stored while the pixel position for storing the processing results in the horizontal direction is within the processing range. .

FIG. 23 shows an example of the configuration of an image data processing apparatus in which the eighth embodiment is realized by a microcomputer, and those having the same numbers as those in FIG. 2 are the same.

Reference numeral 13 denotes an image line memory which is used as an auxiliary memory for temporarily storing a value obtained by passing one line of the luminance value from the image memory 8 in the horizontal direction through the low-pass filter. The image line memory 10 shown in FIG. An image line memory for storing passed values has been added.

In the video signal processing apparatus configured as described above, in addition to accessing the luminance value from the image line memory 13, the edge start point detection means 2, the edge end point detection means 3, the edge start point color difference value The operations of the correcting means 4, the edge end point color difference value correcting means 5, and the vertical color difference value correcting means 6 are the same as in the first embodiment.

Table 4 shows numerical data obtained as a result of correcting the horizontal color difference value of the luminance signal obtained by subjecting the luminance signal and the color difference signal of the original image to low-pass filtering. In the present embodiment, the color difference value is corrected at Th1 = α = β = 5. Each row indicates a pixel position,
Column A shows the number of the pixel position, Column B shows the luminance value of the original image,
Column C is a color difference value of the original image, column D is a color difference value when the color difference value is corrected according to the first embodiment using the luminance value of the original image, and column E is a low-pass filter for the luminance value of the original image. The luminance values after the processing and the F column indicate the color difference values of the image that has been subjected to the same processing as in the first embodiment using the luminance values after the low-pass filter processing. The G column shows the relationship between the pixels detected at that time. However, the color difference value is R-
There are two signals, Y and BY signals, but the processing is the same, so only the RY signal will be shown.

[0264]

[Table 4]

FIG. 24 is a graph showing the result of the processing in (Table 4). The horizontal axis indicates the pixel position, and the vertical axis indicates the luminance value or color difference value of each pixel.

FIG. 24A shows the luminance values of the original image, and FIG. 24B shows the color difference values of the original image.
FIG. 24C shows the color difference values obtained as a result of performing the same processing as in the first embodiment based on the luminance values of the original image. FIG.
4 (d) shows the luminance value after performing the low-pass filter processing on the luminance value of the original image. n1 indicates a luminance edge start point, n2 indicates a luminance edge end point, and n2 + b
Indicates the end point of the luminance smoothing portion.

FIG. 24E shows the color difference value as a result of correcting the color difference value according to the eighth embodiment, where k2 indicates the end point of the color fringing portion of the color difference value. This is the result of performing the same processing as in the first embodiment.

When the luminance values of the original image in Table 4 are used,
The edge start point is detected at the second pixel, and the edge end point is detected at the fifth pixel. Since the luminance value of the edge end point is 146, the condition of the luminance value smoothing region is that the luminance value is 141 or more and 1
It becomes a continuous pixel having a value of 51 or less.

In the smooth region satisfying the condition from the edge end point n2, the luminance value becomes 140 at the ninth pixel, deviating from the condition satisfying the smooth region. Therefore, the smooth region is up to the eighth pixel.

Looking at the luminance value of the ninth pixel, FIG.
As is clear from (a), the condition deviates from the condition for filling the smooth area by one pixel. This pixel is considered to be due to the influence of noise.

This is because the vicinity is a portion that can be visually regarded as a uniform area. Even if the luminance value is shifted by one pixel in the vertical direction from the ninth pixel, the luminance value is similar to that of the ninth pixel. It can be determined that there is no pixel having, and this is due to the influence of noise.

In order to correct the color difference value, when the color difference value is examined in the range of the luminance smooth region from the edge end point to the eighth pixel, the color difference value continues to increase, and the fluctuation of the color difference value is small. The pixel cannot be detected.

As a result, as shown in FIG. 24C, the same color difference values as in FIG. 24B are output without correcting the color difference values. Of course, the image quality is not improved. In other words, it can be seen that since the luminance includes noise, the color difference value cannot be corrected because the luminance smooth region was not correctly detected.

Therefore, in the present embodiment, the color difference value is corrected by using the luminance value obtained by reducing the noise through the low-pass filter processing means 12 for the luminance value.

The luminance values when the luminance values have passed through the low-pass filter processing means 12 are shown in column E of (Table 4), and the characteristic diagram is shown in FIG. The noise of the luminance value that was present at the ninth pixel of the original image is eliminated, and a uniform area is formed. At this time, with respect to the luminance value passed through the low-pass filter processing means 12, the pixel positions of the edge start point and the edge end point are the same as the case of the luminance value of the original image.

The luminance value at the end point of the edge is 146, and the condition for obtaining a smooth region of the luminance value is a continuous pixel whose luminance value falls within a range of 141 to 151. Therefore, the 15th pixel from the edge end point is a luminance smooth region. That is,
The color difference values of the original image from the edge end point to the fifteenth pixel are continuously increasing from the edge end point.
The pixel position k2 where the monotone increase of the color difference value ends at the pixel can be detected. Next, the color difference value from the edge end point to k2 is corrected using the average value of the color difference values from the pixel position k2 to the fifteenth pixel. The results are shown in column F of (Table 4) and FIG. 24 (e).

On the other hand, when the light did not pass through the low-pass filter processing means 12, the color difference value was not changed and the color blur remained, but the luminance value of the original image was passed through the low-pass filter processing means 12. This can reduce color blur and improve image quality. Thereafter, by executing the processing by the vertical color difference value correcting means 6, the image quality is further improved as compared with the first embodiment.

In each of the above embodiments, the image memory is used as the storage means. However, the present invention is not limited to this. Of course, other storage means such as a magnetic disk, an optical disk, and an external RAM may be used.

In each of the above embodiments, the video signal is a luminance signal and color difference signals RY and BY. However, the present invention is not limited to this, and the video signal is composed of a luminance signal and a color difference signal. For example, in the case of the PAL signal, the transmitted color difference signals RY and BY are stored in the image data storage means, and in the case of the NTSC signal, the transmitted color difference signals I and Q are stored in the color difference signal. After being converted into signals RY and BY, the signals can be stored in the image data storage means. Further, when the image signal is composed of an RGB signal, the image signal may be converted into a luminance signal and a color difference signal, then subjected to the color difference value correction according to the present invention, and then converted into an RGB signal and output.

In each of the above embodiments, the color difference signal R-
Although the configuration is such that both Y and BY are corrected, the present invention is not limited to this, and the color difference signal RY in which color blur in a still image is particularly conspicuous.
Of course, it is also possible to adopt a configuration in which only the correction is made. in this case,
Since the processing time is reduced by half and the processing result can be output faster, it is particularly effective when the video signal is continuously input and stored.

In each of the above embodiments, the color difference values of all the pixels in the color difference value change area are corrected. However, the present invention is not limited to this, and a part of them may be corrected.

In each of the above embodiments, the color difference value is corrected on both the edge start point side and the edge end point side. However, the present invention is not limited to this, and only one of them may be corrected.

In the eighth embodiment, the low-pass filter processing is always performed. However, the present invention is not limited to this. For example, the low-pass filter processing is performed only when the color difference value cannot be corrected in the luminance value smooth region. The configuration may be such that the processing is executed and the color difference value is corrected again.

[0284]

As is apparent from the above description, the present invention has an advantage that the degree of correction of color bleeding of an image can be improved and image quality deterioration at the time of correction can be suppressed.

[Brief description of the drawings]

FIG. 1 is a block diagram of a video signal correction device according to a first embodiment of the present invention.

FIG. 2 is a configuration diagram when the video signal correction device of the embodiment is realized by a microcomputer.

FIG. 3 is a diagram illustrating pixel positions of an image memory according to the embodiment.

FIG. 4 is an overall flowchart for explaining an operation in the embodiment.

FIG. 5 is an overall flowchart for explaining an operation of correcting a color difference value in a horizontal direction of an image according to the embodiment.

FIG. 6 is a flowchart for detecting an edge start point in the embodiment.

FIG. 7 is a flowchart for detecting an edge end point in the embodiment.

FIG. 8 is a flowchart of a process for correcting a color difference value from an edge start point in the embodiment.

FIG. 9 is a flowchart of a process for correcting a color difference value from an edge end point in the embodiment.

FIG. 10 is a flowchart of a process for correcting a color difference value in a vertical direction of an image according to the embodiment.

11A is a characteristic diagram of a luminance value when one line in the horizontal direction of the original image is viewed, and FIG. 11B is a diagram illustrating a color difference value of a pixel corresponding to (a) of the original image. Characteristic diagram, FIG.
Is a characteristic diagram when the color difference value in the embodiment is corrected,
FIG. 14D is a characteristic diagram of the color difference values in the second embodiment, which are corrected so that the color difference values are monotonically increased.

12A is a characteristic diagram of luminance values of an original image, FIG. 12B is a characteristic diagram of color difference values corresponding to FIG. 12A of the original image, and FIG. FIG. 4D is a characteristic diagram of the color difference values obtained by performing the correction in the horizontal direction and then performing the correction in the vertical direction after performing the correction in the horizontal direction.

FIG. 13 is a block diagram of a video signal correction device according to a third embodiment of the present invention.

FIG. 14 is a diagram illustrating pixel positions and directions of an image memory according to the same embodiment.

FIG. 15 is an overall flowchart for explaining an operation of correcting a color difference value of an image in a horizontal direction according to the embodiment.

16A is a characteristic diagram of a luminance value of an original image, FIG. 16B is a characteristic diagram of a color difference value of the original image, and FIG.
FIG. 13D is a characteristic diagram of the color difference values after the correction in the horizontal forward direction according to the third embodiment, and FIG. 14D is a characteristic diagram of the color difference values after the correction in the horizontal reverse direction according to the third embodiment; FIG. 11E is a characteristic diagram of the color difference value after the correction in the horizontal forward direction in the fourth embodiment, and FIG. 11F is a characteristic diagram of the color difference value after the correction in the horizontal reverse direction in the fourth embodiment. It is.

FIG. 17 is a block diagram of a video signal correction device according to a fifth embodiment of the present invention.

FIG. 18 is an overall flowchart illustrating an operation of correcting a color difference value of an image in a horizontal direction according to the embodiment.

19A is a characteristic diagram of a luminance value of an original image, FIG. 19B is a characteristic diagram of a color difference value of the original image, and FIG.
FIG. 14D is a characteristic diagram of color difference values after correction in the horizontal forward direction according to the fifth embodiment. FIG. 14D is a characteristic diagram of color difference values after correction in the horizontal reverse direction according to the fifth embodiment. FIG. 11E is a characteristic diagram of color difference values after correction in the horizontal forward direction in the sixth embodiment, and FIG. 11F is a characteristic diagram of color difference values after correction in the horizontal reverse direction in the sixth embodiment. It is.

FIG. 20 is a flowchart illustrating a process of correcting a color difference value in the vertical direction according to the seventh embodiment.

21A is a characteristic diagram of luminance values of an original image, FIG. 21B is a characteristic diagram of color difference values corresponding to FIG. 21A of the original image,
FIG. 13C is a characteristic diagram of the color difference value when the processing in the horizontal direction is performed, and FIG. 14D is a diagram in which the horizontal direction is corrected using the seventh embodiment, FIG. 9 is a characteristic diagram of a color difference value when the color difference value is corrected.

FIG. 22 is a block diagram of a video signal correction device according to an eighth embodiment of the present invention.

FIG. 23 is a configuration diagram when the video signal correction device of the embodiment is realized by a microcomputer.

24A is a characteristic diagram of luminance values of an original image, FIG. 24B is a characteristic diagram of color difference values of the original image, and FIG.
FIG. 14D is a characteristic diagram of color difference values after correction in the horizontal direction in the first embodiment. FIG. 14D shows luminance values after low-pass filtering is performed on the luminance values of the original image according to the eighth embodiment. FIG. 19E is a characteristic diagram of the color difference values after performing the horizontal correction based on the luminance values after the low-pass filter processing according to the eighth embodiment.

FIG. 25 is a flowchart of a process of correcting a color difference value at a horizontal edge start point of an image according to the first embodiment.

FIG. 26 is a flowchart of a process of correcting a color difference value at a horizontal edge start point of an image according to the first embodiment.

FIG. 27 is a flowchart of a process of correcting a color difference value at an edge end point in the horizontal direction of an image according to the first embodiment.

FIG. 28 is a flowchart of a process of correcting a color difference value at a horizontal edge end point of an image according to the first embodiment.

FIG. 29 is a block diagram of a conventional video signal correction device.

[Explanation of symbols]

 1, 8 Image memory 2 Edge start point detection means 3 Edge end point detection means 4 Edge start point color difference value correction means 5 Edge end point color difference value correction means 6 Vertical color difference value correction means 9 Microcomputer 10, 13 Image line memory 11 Access direction switching Device 12 low-pass filter processing means 103 contour correction circuit

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H04N 9/43-11/22

Claims (20)

(57) [Claims] 1. A storage unit for storing a luminance value and a color difference value of each pixel of an image, and a luminance signal generated from the stored luminance value is input, and the luminance value in a predetermined direction of the image is a predetermined value. A brightness change position detecting means for detecting a position of a pixel at at least one end of the brightness change region that increases or decreases; and the brightness value from the detected position of the pixel at the end toward the outside of the brightness change region. A stable brightness region detecting means for detecting a stable brightness region in which the change of the brightness is equal to or less than a predetermined brightness value; and wherein the color difference value in the detected brightness stable region in a predetermined direction of the image increases or decreases by a predetermined value or more. A color difference value change area is detected, and at least a part of the color difference values stored in the storage means in the detected color difference value change area is converted to a color difference value with a small change in the color difference value outside the color difference value change area. A color difference value correction unit that corrects the color difference value using the color difference value of the fixed area, a plurality of the color difference values that are continuous in a direction perpendicular to the predetermined direction of the corrected image, the luminance value is within a predetermined value range, A video signal correction device comprising: a vertical color difference value correcting unit that selects a plurality of pixels and smoothes the color difference values of those pixels. 2. The vertical chrominance value correction means reads out luminance values and chrominance values of at least three pixels continuous from the storage means perpendicular to the predetermined direction. Selecting upper and lower pixels that are perpendicular to the predetermined direction with respect to the target pixel satisfying a condition of −δ or more and Yh + δ or less, and changing a color difference value of the target pixel using the color difference value of the selected pixel; The video signal correction device according to claim 1, wherein: 3. A vertical color difference value correcting means, comprising: at least three consecutive vertical color difference values from a storage means in which an image having a color difference value corrected in the predetermined direction is stored.
The luminance value and the color difference value of the pixel or more are read, and the absolute value of the difference between the luminance values is Y% of Yh with respect to the luminance value Yh of the target pixel.
The method further comprises selecting upper and lower pixels that are not more than (5 ≦ Y ≦ 30) in the vertical direction of the pixel of interest and changing the color difference value of the pixel of interest using the color difference value of the selected pixel. Item 3. The video signal correction device according to Item 1. 4. The image processing apparatus further comprises low-pass filter processing means for sequentially taking out luminance values of pixels continuous in the predetermined direction of the image from the storage means and passing the taken-out luminance values through a low-pass filter, wherein the luminance change position detecting means comprises: 2. The video signal correction device according to claim 1, wherein a luminance value processed by said low-pass filter processing means is used. 5. The color difference value correcting means according to claim 1, wherein the color difference value is changed using a representative value C1 of the color difference values in a continuous color difference value stable region several pixels from the luminance change region. The video signal correction device according to the above. 6. The video signal correction apparatus according to claim 1, wherein the vertical chrominance value correction unit refers to upper and lower two pixels that are continuous in a vertical direction of the image with respect to the target pixel. 7. The luminance change position detecting means sequentially retrieves the luminance values of the pixels consecutive in the predetermined direction from the storage means, and calculates n1 at which the luminance value starts to increase or decrease by a predetermined value or more.
Edge start point detection means for detecting a pixel as an edge start point, and edge end point detection means for setting an n2 (≧ n1 + 2) th pixel as an edge end point in which the luminance value continuously increases or decreases by a predetermined value or more from the edge start point 2. The video signal correction device according to claim 1, comprising: 8. The color difference value correction means sequentially extracts color difference values corresponding to pixels from the (n1-a) th (<n1) th pixel to the edge start point in the luminance stable area, and calculates k1 (>
Edge start point color difference value correction for changing at least a part of the color difference value of a pixel whose color difference value continuously increases or decreases by a predetermined value or more from the (n1-a, ≤n1-2) th pixel to the edge start point. Means, and n2 in the luminance value stable area from the pixel at the edge end point.
The color difference values corresponding to each pixel up to + b (> n2) are sequentially extracted, and k2 (≧ n2 + 2, <
An edge end point color difference value correcting means for changing at least a part of the color difference value of a pixel whose color difference value continuously increases or decreases by two or more pixels up to an (n2 + b) th pixel. 8. The video signal correction device according to 7. 9. A chrominance value correction unit sequentially extracts chrominance values corresponding to pixels from the pixel at the edge end point to the n2 + b (> n2) th pixel in the luminance stable area, and obtains k2 ( ≧ n2 + 2, <n2 + b) edge end point color difference value correction means for changing at least a part of the color difference value of a pixel whose color difference value continuously increases or decreases by a predetermined value or more up to two or more pixels up to the (n) + th pixel; The value change is performed by correcting the color difference value while sequentially taking out the luminance value and the color difference value of the pixel continuous in the predetermined direction of the image, and then adjusting the luminance of the pixel continuous in the direction opposite to the predetermined direction of the corrected image. The video signal correction apparatus according to claim 7, wherein the correction is performed by re-correcting the color difference value while extracting the value and the color difference value. 10. A luminance change position detecting means sequentially retrieves luminance values of pixels continuous in a predetermined direction of the image from the storage means, and two or more pixels successively start increasing or decreasing the luminance value by a predetermined value or more. An edge start point detection unit having an edge start point of the ith pixel, wherein the chrominance value correction unit includes a chrominance value corresponding to each pixel from a pixel in an n1-a (<n1) th luminance stable area to the edge start point. Are sequentially extracted and k1 (> n1-a, ≤n
1-2) an edge start point color difference value correction unit that changes at least a part of the color difference values of pixels whose color difference values continuously increase or decrease by a predetermined value or more from the second pixel to the edge start point by two or more pixels, The change of the chrominance value is performed by correcting the chrominance value while sequentially extracting the luminance value and the chrominance value of the pixel continuous in a predetermined direction of the image, and then correcting the pixel of the corrected image in the direction opposite to the predetermined direction. 2. The video signal correction device according to claim 1, wherein the correction is performed by re-correcting the color difference value while extracting the luminance value and the color difference value. 11. The color difference value correcting means for changing a color difference value using a representative value C1 of a color difference value of each pixel in the color difference value stable area.
The video signal correction device according to the above. 12. The color difference value correction device according to claim 8, wherein the edge end point color difference value correction means changes the color difference value using a representative value C2 of the color difference value of each pixel in the color difference value stable area. Video signal correction device. 13. The color difference value correcting means for an edge start point includes at least a representative value C1 of the color difference values of the respective pixels in the k1 to k1-c (≧ n1-a) th color difference value stable areas in which the change in the color difference values is small. 11. The video signal correction device according to claim 8, wherein a color difference value of each pixel from a k1st pixel to the edge start point is changed. 14. An edge start point color difference value correction means uses a representative value C1 of color difference values of each pixel in a k1 to k1-c (≧ n1-a) th color difference value stable region with little change in color difference value. To change the color difference value between the k1st pixel and the n1-e (<n1,> k1) th pixel,
-The color difference value of a pixel from the e-th pixel to the edge start point is changed from the color difference value of the edge start point to the representative value C1 to a continuously increased or decreased value. 11. The video signal correction device according to 8 or 10. 15. The edge end point color difference value correction means uses a representative value C2 of a color difference value of each pixel in a k2 to k2 + d (≦ n2 + b) -th color difference value stable area having a small change in the color difference value to at least the edge. 10. The video signal correction device according to claim 8, wherein a color difference value of a pixel from an end point to a k2th pixel is changed. 16. The edge end point color difference value correction means uses the representative value C2 of the color difference value of each pixel in the k2 to k2 + d (≦ n2 + b) th color difference value stable regions with little change in the color difference value, and calculates the k2th color difference value. And the color difference value between the pixel of nth and the n2 + e (> n2, <k2) th pixel is changed, and
The color difference value of the pixel from the pixel to the end point of the edge,
10. The video signal correction device according to claim 8, wherein the color difference value at the edge end point is changed to a value that is monotonically increased or monotonously decreased from the representative value C2. 17. The chrominance value correction means uses the representative value C1 of the chrominance value of each pixel in the k1 to k1-c (≧ n1-a) th chrominance value stable area with little variation in chrominance value to calculate k1 An edge start point color difference value correction means for changing the color difference value of each pixel from the pixel to the edge start point; and a k2 to k2 + d (≦ n2 + b) th color difference value stable area with little change in the color difference value. Edge end point color difference value correction means for changing the color difference value of the pixel from the edge end point to the k2th pixel using the representative value C2 of the pixel color difference value; and the color difference value from the edge start point to the edge end point. Is the representative value C
8. The video signal correction apparatus according to claim 7, further comprising: a luminance-change-portion color difference value correction unit that continuously changes from 1 to a representative value C2 to a monotonically increasing or monotonically decreasing value. 18. A storage means for storing a luminance value and a chrominance value of each pixel of an image, and sequentially retrieves luminance values of pixels continuous in a predetermined direction of the image from the storage means. Starting point detecting means for detecting an n1th pixel which starts increasing or decreasing as an edge starting point, and an edge having an n2 (≧ n1 + 2) th pixel at which the continuous increase or decrease of the luminance value from the edge starting point ends as an edge end point End point detecting means; and sequentially extracting color difference values corresponding to each pixel in the luminance stable area from the n1-a (<n1) th pixel having a small variation in luminance value to the edge start point, and obtaining k1 (> n1-a, ≤ n1-
2) an edge start point color difference value correction means for changing at least a part of the color difference value of a pixel whose color difference value continuously increases or decreases by two or more pixels from a pixel to the edge start point; and the edge in a luminance stable area. From the pixel at the end point, n2 +
The color difference values corresponding to each pixel up to the b (> n2) th pixel are sequentially extracted, and k2 (≧ n2 +
Edge end point color difference value correction means for changing at least a part of the color difference value of a pixel in which the color difference value continuously increases or decreases by two or more pixels up to the (<n2 + b) th pixel; and having the corrected color difference value. From the storage unit in which the image is stored, the luminance value and the color difference value of at least three pixels continuous in the vertical direction with respect to the predetermined direction are read, and the luminance value of the pixel of interest that falls within a predetermined value range of the luminance value of the pixel of interest is read A video signal correcting device comprising: a vertical color difference value correcting unit that selects upper and lower pixels that are continuous in the vertical direction, and changes a color difference value of the target pixel using the color difference value of the selected pixel. . 19. A storage unit for storing a luminance value and a color difference value of each pixel of an image, an edge start point detection unit for detecting an edge start point of the image, and an edge start point color difference for changing a color difference value with respect to the edge start point. Value correction means, and vertical color difference value correction means for changing the color difference value in the vertical direction of the image, wherein the edge start point detection means sequentially retrieves the luminance values of pixels continuous in a predetermined direction of the image from the storage means. ,
The luminance value starts to increase or decrease continuously for two or more pixels n
The first pixel is set as an edge start point, and the color difference value correction means sequentially extracts color difference values corresponding to each pixel from the n1-a (<n1) th pixel having a small variation in luminance value to the edge start point. , K1
Changing at least a part of the color difference value of a pixel whose color difference value continuously increases or decreases by two or more pixels from the (> n1-a, ≤n1-2) th pixel to the edge start point; While taking out the color difference value of the image in the direction opposite to the predetermined direction, the color difference value is corrected again by using the edge start point detection means and the edge start point portion color difference value correction means, and then the vertical color difference value correction means A luminance value and a color difference value of at least three pixels continuous in a direction perpendicular to the predetermined direction are read from the storage unit in which the image having the corrected color difference value is stored, and the luminance value of the pixel of interest is a predetermined value. And selecting the upper and lower pixels that are consecutive in the vertical direction of the pixel of interest within the range of, and changing the color difference value of the pixel of interest using the color difference value of the selected pixel. Signal correction device. 20. A storage unit for storing a luminance value and a color difference value of each pixel of an image, an edge start point detection unit for detecting an edge start point of the image, an edge end point detection unit for detecting an edge end point of the image, An edge end point color difference value correction unit that changes a color difference value with respect to the edge end point portion, and a vertical color difference value correction unit that changes a color difference value in a vertical direction of the image, wherein the edge start point detection unit outputs the color difference value from the storage unit. The luminance values of pixels consecutive in a predetermined direction of the image are sequentially extracted,
The luminance value starts to increase or decrease continuously for two or more pixels n
The first pixel is set as the edge start point, and the edge end point detection means ends the increase or decrease of the continuous brightness value from the edge end point n2 (≧ n1 +
2) The pixel at the edge is defined as the edge end point, and the edge end point color difference value correcting means starts n2 + b (> n
2) The color difference values corresponding to the respective pixels up to the 画素 th pixel are sequentially extracted, and k2 (≧ n2 + 2, <n2) from the edge end point.
+ B) changing at least a part of the color difference value of the pixel in which the color difference value continuously increases or decreases by two or more pixels up to the (th) pixel, and thereafter changes the color difference value of the corrected image in a direction opposite to the predetermined direction. While taking out, the edge start point detecting means,
The color difference value is corrected again by using the edge end point detection means and the edge end point color difference value correction means, and thereafter, the storage means in which an image having the color difference value corrected again by the vertical color difference value correction means is stored. And reads a luminance value and a chrominance value of at least three or more pixels which are continuous in the vertical direction with respect to the predetermined direction, and the luminance value of the target pixel is continuous in the vertical direction of the target pixel falling within a predetermined value range A video signal correction apparatus, wherein upper and lower pixels are selected, and a color difference value of the target pixel is changed using a color difference value of the selected pixel.