JP5067044B2 - Image processing apparatus and image processing method - Google Patents

Description

  The present invention relates to an image processing apparatus and an image processing method.

  In analog NTSC broadcasting, a composite video signal is used. When YC separation is performed to separate the composite video signal into two systems of luminance signal Y and color signal C, the color signal amplitude-modulated by the color subcarrier signal is luminance. If it leaks into the signal, dot interference occurs.

  On the other hand, in HDTV (high definition television broadcasting) in terrestrial digital broadcasting, a component video signal in which a luminance signal and a color signal are separated is employed. For this reason, when broadcasting video content using the composite video signal system as a source, it is necessary to perform YC separation on the broadcast station side. At this time, if the YC separation performance is low, a lot of dot interference is included in the received video of the terrestrial digital HD broadcast.

  Also, composite video signal video has an aspect ratio of 4 to 3 and 525 scanning lines, whereas HD broadcast video has an aspect ratio of 16 to 9 and 1125 scanning lines. is there. Therefore, when broadcasting video content using the composite video signal system as a source, it is necessary to perform scaling processing on the broadcasting station side.

  As described above, the spatial characteristics of dot interference included in HD broadcasting are greatly different from those of conventional analog broadcasting. In addition, the terrestrial digital HD broadcast has a problem that the above-described dot disturbance phenomenon occurs very often.

  Patent Documents 1 to 3 disclose detection and improvement methods for dot interference.

JP-A-2005-328150 JP 2003-116150 A JP 2003-116151 A

  The technique of Patent Document 1 is for an HD video in which an SD signal is up-converted and a side panel is added, and is a method for improving dot interference at this time. According to Patent Document 1, first, an up-converted video is determined using video format information and side panel additional information, and the up-converted video is down-converted on the receiver side. Then, it is converted into a composite video signal, and YC separation and scaling are performed again to improve dot interference.

  However, in the method of Patent Document 1, even if the above processing itself is not performed or the processing is performed on a video in which a video having HD resolution and an up-converted video including dot interference are mixed. Since the filter is applied to all pixels, the resolution can be degraded.

  The techniques of Patent Documents 2 and 3 are methods for detecting and improving dot interference that occurs when YC separation is performed on the SD composite video signal on the receiver side. However, it is difficult to detect the dot interference included in the HD video described above by the methods of Patent Documents 2 and 3.

  Therefore, the present invention has been made in view of the above problems, and an object of the present invention is to detect a dot interference position even for an input of a video signal of a high-definition television broadcast, and to detect a detection position. It is an object of the present invention to provide a new and improved image processing apparatus and image processing method capable of obtaining a better video without degrading the overall resolution by performing improvement processing on the image.

  In order to solve the above problems, according to an aspect of the present invention, a pattern corresponding to noise in a video signal of high-definition television broadcasting including noise generated by separating a video signal into two or more component signals. An image processing apparatus comprising: a detection unit that detects a generation position of an image; and an improvement processing unit that reduces noise at a generation position of a pattern detected by the detection unit in a video signal of a high-definition television broadcast Is provided.

  With this configuration, the detection unit detects the occurrence position of a pattern corresponding to noise in a video signal of high-definition television broadcasting that includes noise generated by separating the video signal into two or more component signals, and performs an improvement process. The unit reduces the noise at the pattern generation position detected by the detection unit in the video signal of the high-definition television broadcast. Here, the pattern corresponding to noise is, for example, dot interference.

  The video signal is configured for each frame, and the detection unit detects a color edge generation position where the frequency value of the color signal of the component signal is higher than a predetermined threshold, and the luminance of the component signal. A first detection unit that detects a characteristic change position having a frequency value within a predetermined range for the signal, and a phase inversion occurs at a position that is separated from the luminance signal by a predetermined direction in the frame with respect to the luminance signal. And a second detection unit for detecting a position to detect the pattern generation position based on a detection result of at least one of the color edge detection unit, the first detection unit, and the second detection unit. It may be.

  With this configuration, for the color signal of the component signals, the color edge detection unit detects the occurrence position of the color edge whose frequency value is higher than a predetermined threshold, and for the luminance signal of the component signals, the first detection unit. Detects a characteristic change position where the frequency value is within a predetermined range, and the second detection unit reverses the phase with respect to a position where the luminance signal of the component signals is separated by a predetermined interval in the frame in a predetermined direction. The detection position is detected, and the detection unit detects a generation position of a pattern corresponding to noise based on a detection result of at least one of the color edge detection unit, the first detection unit, and the second detection unit. . The predetermined direction is, for example, the vertical direction of the frame, and the predetermined spaced position is, for example, several lines above or below.

  The video signal is configured for each frame, and the detection unit includes a third detection unit that detects a position where a phase inversion occurs in the luminance signal among the component signals between two frames acquired in time series. And a pattern generation position may be detected based on the detection result of the third detection unit. With this configuration, the third detection unit detects a position where the phase inversion occurs with respect to the luminance signal among the component signals between two frames acquired in time series. Based on the detection result, the pattern generation position is detected.

  The improvement processing unit may perform an average process of the detected position of the pattern and the output value of the position adjacent in the predetermined direction in the frame. With this configuration, the noise can be reduced by averaging the output values at positions adjacent to each other in a predetermined direction in the frame for the pattern generation position detected by the detection unit. The predetermined direction is, for example, the vertical direction of the frame.

  The improvement processing unit may perform an average process of output values of pattern generation positions between two frames acquired in time series. With this configuration, the noise can be reduced by averaging the output values of the pattern generation positions between two frames acquired in time series for the pattern generation positions detected by the detection unit.

  In order to solve the above problem, according to another aspect of the present invention, a high-definition television broadcast video signal including noise generated by separating a video signal into two or more component signals is converted into noise. An image processing method comprising: detecting a corresponding pattern generation position; and reducing noise with respect to the pattern generation position detected in the step of the high-definition television broadcast video signal. Provided.

  According to the present invention, it is possible to detect a dot interfering position even when an image signal of a high-definition television broadcast is input, and to improve the detected position without degrading the overall resolution. Can be obtained.

  Exemplary embodiments of the present invention will be described below in detail with reference to the accompanying drawings. In addition, in this specification and drawing, about the component which has the substantially same function structure, duplication description is abbreviate | omitted by attaching | subjecting the same code | symbol.

(First embodiment)
First, a first embodiment of the present invention will be described. FIG. 1 is a block diagram showing a first embodiment of the present invention.

  The dot interference signal processing unit 100 includes a dot interference detection unit 110, a dot interference improvement processing unit 120, and the like. The dot interference detection unit 110 receives a high definition television broadcast (HDTV) video signal and detects a dot interference occurrence position. The dot interference improvement processing unit 120 receives a high-definition television broadcast video signal and a detection result indicating a dot interference occurrence position in the dot interference detection unit 110 and performs video signal improvement processing on the dot interference occurrence position. And output an improved video signal.

  Here, the video signal of the high-precision television broadcast (HDTV) is, for example, an interlaced YUV422 format. The video signal of the present invention is not limited to this example. The HDTV video signal input to the dot interference signal processing unit 100 includes dot interference, which is noise generated by performing YC separation processing for separating the video signal into two systems of the luminance signal Y and the color signal C. . Note that dot interference is an example of a pattern corresponding to noise.

  For example, when broadcasting a video content using a composite video signal system as a source, YC separation is performed on the broadcasting station side, and therefore, HDTV video signals may contain dot interference. The composite video signal video has an aspect ratio of 4 to 3 and the number of scanning lines of 525, whereas the HDTV video has an aspect ratio of 16 to 9 and the number of scanning lines of 1125. . Therefore, in HDTV, when video content using a composite video signal system is broadcast as a source, scaling processing is performed on the broadcast station side.

  Next, with reference to FIG. 2, each component of this embodiment is demonstrated. FIG. 2 is a block diagram showing this embodiment.

  The dot interference signal processing unit 100 receives a video signal from the video decoder unit 102. The video decoder unit 102 decodes the received video signal of the terrestrial digital HD broadcast, and outputs the separated luminance signal Y and color signal C to the dot interference signal processing unit 100.

  The dot disturbance detection unit 110 is an example of a detection unit, and includes a color edge detection unit 112, a horizontal waveform detection unit 114, a vertical direction phase inversion detection unit 116, and an addition unit 118.

  The color edge detection unit 112 detects the occurrence position of a color edge whose frequency value is higher than a predetermined threshold for the color signal among the component signals. The color edge detection unit 112 receives a color signal from the video decoder unit 102. The color signal is composed of a U signal and a V signal. The color edge detection unit 112 multiplies the output value of the U signal and the output value of the V signal of the target pixel by a filter coefficient. Also, the output values of the U signal and the V signal are multiplied by filter coefficients for a plurality of pixels in the horizontal direction near the target pixel with the target pixel as the center. Then, all the multiplication results of a plurality of pixels centered on the target pixel are summed, and the filtering value, which is the sum, is compared with an arbitrary threshold value Th1. When the filtering value is greater than the threshold value, it is determined that a color edge has been detected for the target pixel. A high-pass filter can be applied as the filter.

  In the case of YC separation, when the frequency of the original color signal is high, the dot signal is generated by the above-described configuration based on the cause of dot interference that the color signal amplitude-modulated by the color subcarrier signal is likely to leak into the luminance signal. It is possible to detect a target pixel in which interference may occur.

  The color edge detection unit 112 will be described with reference to FIG. FIG. 3 is an explanatory diagram showing an array of pixels for explaining pixels in which a color edge is detected.

  The target pixel to be detected by the color edge detection unit 112 is j lines in the vertical direction and i-th pixel in the horizontal direction. In the case of color edge detection in the U signal, the output value of the U signal at that time is U [i] [j], and the filter coefficient is filter [k]. When using two pixels on the left and right in the vicinity centering on the target pixel for the color edge detection of the target pixel, five pixels are necessary for the color edge detection. The filter coefficient k is assumed to be k = 0, 1, 2, 3, 4. At this time, the filtering value can be expressed by Equation 1 below.

  When the filtering value Fu [i] [j] exceeds the threshold Th1, it is determined that the color edge is detected in the target pixel. The threshold value Th1 used in the color edge detection unit 112 is arbitrarily set according to the input signal and the required detection accuracy.

  The horizontal waveform detection unit 114 is an example of a first detection unit, and detects a waveform characteristic change position whose frequency value is within a predetermined range for a luminance signal among component signals. The horizontal waveform detection unit 114 receives the luminance signal from the video decoder unit 102. The horizontal waveform detection unit 114 multiplies the output value of the luminance signal of the target pixel by a filter coefficient. Also, for each of a plurality of pixels in the horizontal direction near the target pixel with the target pixel at the center, each output value of the luminance signal is multiplied by a filter coefficient. Then, all the multiplication results of a plurality of pixels centered on the target pixel are summed, and the filtering value, which is the sum, is compared with an arbitrary threshold Th2. When the filtering value is larger than the threshold value, it is determined that the waveform characteristic change is detected for the target pixel.

  As the filter, for example, a band-pass filter that passes frequencies in the vicinity of 3.58 / (1125/525) MHz can be applied. This is because the spatial horizontal characteristics of dot interference change because the up-converted video of HD broadcasting is scaled approximately 1125/525 times for the color subcarrier signal frequency of 3.58 MHz. It is based on.

  The target pixel to be detected by the horizontal waveform detection unit 114 is the j-th pixel in the vertical direction and the i-th pixel in the horizontal direction. The output value of the luminance signal at that time is Y [i] [j], and the filter coefficient is filter2 [k]. When using two pixels on the left and right adjacent to the target pixel for the horizontal waveform detection of the target pixel, five pixels are required for the horizontal waveform detection. By replacing U in U [i] [j] in FIG. 3 describing the color edge detection unit 112 with Y, the pixels detected by the horizontal waveform detection unit 114 can be described. The filter coefficient k is assumed to be k = 0, 1, 2, 3, 4. At this time, the filtering value can be expressed by Equation 2 below.

  When the filtering value Fy [i] [j] exceeds the threshold Th2, it is determined that the waveform change characteristic of the horizontal waveform is detected in the target pixel. The threshold value Th2 used in the horizontal waveform detection unit 114 is arbitrarily set according to the input signal and the required detection accuracy.

  The vertical direction phase inversion detection unit 116 is a position of the second detection unit, and detects a position where a phase inversion occurs with respect to a position of the luminance signal separated from the luminance signal in a predetermined direction in a predetermined direction. . The vertical direction phase inversion detection unit 116 sets a plurality of neighboring horizontal pixels centering on the target pixel as a target block, and the target block including the target pixel has no correlation with a block of an arbitrary line. Or correlate with a block of another line. This is based on the fact that the phase of the color subcarrier signal is inverted for each line and the scaling process is performed. A line refers to a continuous arrangement of pixels in the horizontal direction.

  The vertical direction phase inversion detection unit 116 will be described with reference to FIG. FIG. 4 is an explanatory diagram showing an array of pixels for explaining pixels in which vertical phase inversion is detected.

  For example, the target pixel to be detected by the vertical direction phase inversion detection unit 116 is the j-th pixel in the vertical direction and the i-th pixel in the horizontal direction. The output value of the luminance signal at that time is Y [i] [j]. The target block including the target pixel is not correlated with a block composed of 7 pixels on two lines of the target block, assuming that the target block including the left and right three pixels around the target pixel, that is, 7 pixels. Judge that. An expression for determining that there is no correlation can be expressed by the following Expression 3 as k = 0, 1, 2, 3, 4, 5, 6.

  When Expression 3 is satisfied, it is determined that the blocks on the two lines of the target block and the target block have no correlation.

  Furthermore, it is determined that there is a correlation with a block consisting of 7 pixels on 4 lines of the target block. A formula for determining that there is a correlation can be expressed by the following formula 4 as k = 0, 1, 2, 3, 4, 5, 6.

  When Expression 4 is satisfied, it is determined that the target block and the blocks on the four lines of the target block have a correlation. And when satisfy | filling both Numerical formula 3 and Numerical formula 4, a target pixel is detected as a candidate with the possibility that the dot disturbance has generate | occur | produced. Note that the threshold values Th3 and Th4 used in the vertical direction phase inversion detection unit 116 are arbitrarily set according to the input signal and the required detection accuracy.

  Although it is determined that there is no correlation with the block on the second line of the target block and that there is a correlation with the block on the four lines, there is no correlation with the target block or there is a correlation. The determination may be a relationship with a block below the target block.

  In the above detection method, the position of the line for confirming the filter coefficient and the correlation is not fixed but can be arbitrarily changed. Thereby, it is also possible to detect dot interference in accordance with the scaling factor in the scaling process.

  The adder 118 designates the dot interference position by taking (adding) the dot interference candidates in the luminance signal and the color signal.

  The dot disturbance improvement processing unit 120 is an example of an improvement processing unit, and includes a vertical direction improvement processing unit 122. The vertical direction improvement processing unit 122 receives the luminance signal and the color signal from the video decoder unit 102 and the detection result output from the addition unit 118.

  The vertical direction improvement processing unit 122 performs an average process on the target pixel at a position detected as a dot disturbance pattern by the adding unit 118 and a pixel several lines above or several pixels below the target pixel. Do.

  This corresponds to the feature that the phase of dot interference in the terrestrial digital HD broadcasting is inverted every few lines by the scaling process, whereas the phase of dot interference has been inverted every line in the past. It is. Then, by performing the averaging process in the vertical direction, the dot disturbance phase is canceled and the dot disturbance can be improved.

  As described above, according to the dot interference signal processing unit 100 of the present embodiment, first, the video signal of the terrestrial digital HD broadcast received by the video decoder unit 102 is decoded, and the luminance signal and the color signal are output. Next, the dot interference detection unit 110 receives the decoded luminance signal and color signal and detects a dot interference pattern. At that time, for the luminance signal, the horizontal waveform detection unit 114 performs horizontal waveform detection, and the vertical phase inversion detection unit 116 further performs vertical phase inversion detection for the detected target pixel. . As a result, the candidate for the dot interference position is output to the adder 118.

  In addition, for the color signal, the color edge detection unit 112 performs color edge detection and outputs a candidate for the dot interference position to the addition unit 118. Then, the adding unit 118 adds the dot interference candidates in the luminance signal and the color signal, thereby specifying the dot interference position. Then, the dot interference improvement processing unit 120 performs an average process in the vertical direction on the position where the dot interference is detected, improves the dot interference, and outputs the processed video signal to the outside.

  In this embodiment, since the frame memory is not required for detecting the dot disturbance position, the dot disturbance detection circuit can be downsized.

(Second Embodiment)
Next, a second embodiment of the present invention will be described. FIG. 5 is a block diagram showing the present embodiment. The dot interference signal processing unit 200 according to the second embodiment includes a dot interference detection unit 210, a dot interference improvement processing unit 320, a frame memory 230, and the like. The dot disturbance detection unit 210 is an example of a detection unit. The color edge detection unit 112, the horizontal direction waveform detection unit 114, the vertical direction phase inversion detection unit 116, the time direction phase inversion detection unit 117, and the addition Part 118. Since this embodiment is different from the first embodiment in that it includes a time direction phase inversion detection unit 117 and a frame memory 230, the time direction phase inversion detection unit 117 will be described in detail, A description of the components is omitted.

  The time direction phase inversion detection unit 117 sets a plurality of adjacent horizontal pixels centering on the target pixel as a target block. The time direction phase inversion detection unit 117 includes a target block including the target pixel, a block that is higher by an arbitrary line than the target block, or a block that is lower by an arbitrary line than the target block one frame later or one frame before Based on the correlation with the block, it is determined whether the phase of the waveform is inverted at the same position.

  The target pixel to be detected by the time direction phase inversion detection unit 117 is the j-th pixel in the vertical direction and the i-th pixel in the horizontal direction. The output value of the luminance signal at that time is Y [i] [j]. Also, the output value of the luminance signal of the target pixel at the same position after one frame is Y2 [i] [j]. The target block including the target pixel is assumed to be composed of three pixels on the left and right in the vicinity centering on the target pixel, that is, seven pixels.

  A formula for determining that there is a correlation can be expressed by the following formula 5 as k = 0, 1, 2, 3, 4, 5, 6. FIG. 6 shows a state of processing of Formula 5. FIG. 6 is an explanatory diagram showing processing of time direction phase inversion detection.

  When Expression 5 is satisfied, the time direction phase inversion detection unit 117 inverts the phase of the target block including the target pixel and the block two lines above the target block at the same position as the block one frame later. Can be determined. The threshold Th5 used in the time direction phase inversion detection unit 117 is arbitrarily set according to the input signal and the required detection accuracy.

  In the second embodiment, a frame memory 230 is further provided as compared with the first embodiment. The frame memory 230 is used for detection using time axis information in the time direction phase inversion detection unit 117. The frame memory 230 accumulates a plurality of frames in the time axis direction, and outputs a frame before or after the target frame at the time of detection to the time direction phase inversion detection unit 117.

  In the above detection method, the position of the line for confirming the filter coefficient and the correlation is not fixed but can be arbitrarily changed. Thereby, it is also possible to detect dot interference in accordance with the scaling factor in the scaling process.

  As described above, in the present embodiment, the time direction phase inversion detection unit 117 is added to the first embodiment, and the time direction phase inversion detection is performed on the luminance signal. Accordingly, since the time axis information is used in detecting the dot interference position, a frame memory is required. However, since the amount of information is increased as compared with the first embodiment, the dot interference detection accuracy can be improved.

(Third embodiment)
Next, a third embodiment of the present invention will be described. FIG. 7 is a block diagram showing this embodiment.

  The dot interference signal processing unit 300 according to the third embodiment includes a dot interference detection unit 210, a dot interference improvement processing unit 320, a frame memory 230, and the like. The dot disturbance improvement processing unit 320 is an example of an improvement processing unit, and includes a time direction improvement processing unit 322. In the third embodiment, the vertical direction improvement processing unit 122 of the second embodiment is replaced with a time direction improvement processing unit 322.

  The time direction improvement processing unit 322 performs an average process on the target pixel at the position detected as the dot interference pattern by the adding unit 118 and the pixel after one frame or one frame before. The time direction improvement processing unit 322 acquires the luminance signal Y2 and the color signal C2 after one frame or one frame before from the frame memory 230.

  The processing of the time direction improvement processing unit 322 corresponds to the feature that the phase of dot interference is reversed every frame. Then, by performing the averaging process in the time direction, the phase of dot interference can be canceled and dot interference can be improved.

  Note that the time direction improvement processing unit 322 is not limited to the case where the averaging process is performed in the time direction, and noise reduction in the time direction can also be used. The time direction improvement processing unit 322 can obtain a dot interference improvement effect by controlling the reduction rate of noise reduction according to the dot interference detection result.

  As described above, according to each embodiment of the present invention, for example, a YC separation process is performed on the broadcasting station side and a scaling process is performed, a normal HD image, and these images are mixed. It is possible to detect only the position where the dot disturbance occurs even when various types of HD video input such as the one is input. Further, by performing the improvement process on the detection position, it is possible to obtain an image with better image quality without degrading the overall resolution.

  As mentioned above, although preferred embodiment of this invention was described referring an accompanying drawing, it cannot be overemphasized that this invention is not limited to this example. It will be apparent to those skilled in the art that various changes and modifications can be made within the scope of the claims, and these are naturally within the technical scope of the present invention. Understood.

It is a block diagram which shows the 1st Embodiment of this invention. It is a block diagram which shows the same embodiment. It is explanatory drawing which shows the arrangement | sequence of the pixel explaining the pixel from which a color edge is detected. It is explanatory drawing which shows the arrangement | sequence of the pixel explaining the pixel from which a vertical direction phase inversion is detected. It is a block diagram which shows the 2nd Embodiment of this invention. It is explanatory drawing which shows the process of a time direction phase inversion detection. It is a block diagram which shows the 3rd Embodiment of this invention.

Explanation of symbols

100, 200, 300 Dot disturbance signal processing unit 102 Video decoder unit 110, 210 Dot disturbance detection unit 112 Color edge detection unit 114 Horizontal waveform detection unit 116 Vertical direction phase inversion detection unit 117 Time direction phase inversion detection unit 118 Addition unit 120 320 Dot disturbance improvement processing unit 122 Vertical direction improvement processing unit 230 Frame memory 322 Time direction improvement processing unit

Claims (5)

A detection unit for detecting a generation position of a pattern corresponding to the noise of a video signal of high-definition television broadcasting including noise generated by separating the video signal into two or more component signals;
An improvement processing unit that reduces the noise with respect to the generation position of the pattern detected by the detection unit in the video signal of the high-definition television broadcast;
With
The video signal is configured for each frame,
The detection unit includes a third detection unit that detects a position where a phase inversion occurs in the luminance signal among the component signals between the two frames acquired in time series,
An image processing apparatus that detects a generation position of the pattern based on a detection result of the third detection unit . The detector is
A color edge detection unit for detecting a generation position of a color edge having a frequency value higher than a predetermined threshold for the color signal among the component signals;
A first detection unit that detects a characteristic change position in which a frequency value is within a predetermined range for a luminance signal among the component signals;
A second detection unit for detecting a position where a phase inversion occurs with respect to a position separated from the luminance signal in a predetermined direction within the frame in a predetermined direction with respect to the luminance signal;
With
In addition to the detection result of the third detection unit, the generation position of the pattern is detected based on the detection result of at least one of the color edge detection unit, the first detection unit, and the second detection unit. The image processing apparatus according to claim 1. The image processing apparatus according to claim 1 , wherein the improvement processing unit performs an average process of output values of the detected position of the pattern and a position adjacent in a predetermined direction in the frame. The image processing apparatus according to claim 1 , wherein the improvement processing unit performs an average process of output values of the pattern generation positions between the two frames acquired in time series. Detecting a generation position of a pattern corresponding to the noise of a video signal of high-definition television broadcasting including noise generated by separating a video signal configured for each frame into two or more component signals;
Reducing the noise for the occurrence position of the pattern detected in the step of the high-definition television broadcast video signal;
Including
In the detecting step, a position where a phase inversion occurs in the luminance signal among the component signals between the two frames acquired in time series is detected, and the generation position of the pattern is determined based on the detection result. An image processing method to detect .

Images