JP4285396B2 - Block noise reduction device - Google Patents

Description

  The present invention relates to a block noise reduction apparatus, and more particularly to a block noise reduction apparatus that reduces or eliminates block noise generated when encoding and decoding video signals that are digitally compressed and transmitted and recorded.

When digitally compressing a video signal, a method is known in which a plurality of pixels adjacent in both the horizontal and vertical directions are set as one rectangular block and compression encoding is performed in units of the rectangular blocks.
The video signal that has been compression-encoded by such a compression-encoding method is decoded in units of rectangular blocks after passing through a recording medium or a transmission path, and the original video signal is restored.

  In such compression encoding processing and decoding processing, when it is desired to keep the amount of information of the video signal in the recording medium or the transmission path small, the compression rate is set large. A gradation difference occurs. In particular, in a video signal with a gradual gradation change such as a gradation, a gradation difference between adjacent rectangular blocks appears remarkably. In general, noise due to this gradation difference is called block noise and is easily noticeable on the screen.

  Therefore, in order to reduce or eliminate the block noise, a block noise reduction device that corrects a difference in signal level at the boundary between a rectangular block in which block noise is generated and an adjacent block using a filter circuit has been conventionally used. It is known (see, for example, Patent Document 1).

  The conventional block noise reducing device described in Patent Document 1 includes an isolated differential point extracting means for extracting and outputting an isolated differential point from a differential signal obtained by performing differential processing on an input video signal, and a filter for the isolated differential point. Filter means for correcting the signal level difference at the boundary between the rectangular block where the block noise is generated and the adjacent block by processing, and outputting a correction signal; delay means for delaying the input video signal for a predetermined time; An addition means for adding the correction signal and the delayed video signal from the delay means is provided, and a video signal with reduced block noise is extracted from the addition means.

JP 2001-119695 A

  However, in the above conventional block noise reduction device, block noise is reduced by the above filter circuit for a video signal having an image portion whose gradation change is gradual as shown in FIG. 7A. Alternatively, when the removal process is performed, a filter may not be applied to the central portion of the central rectangular block between the block boundary lines I and II, as indicated by 100 in FIG. However, there is a problem that the image appears as vertical stripes on the screen and appears as horizontal stripes when processed in the vertical direction.

In addition, when block noise reduction processing is performed such that a filter is applied to the center of the central rectangular block, there is a problem that high frequency components deteriorate and blurring occurs in the image after the reduction processing.
Furthermore, there are cases where the steps at both ends of the rectangular block are video signals that are not block noise. However, if block noise reduction processing is performed on such video signals, the image quality will be significantly degraded.

  The present invention has been made in view of the above points, and provides a block noise reduction apparatus capable of sufficiently reducing block noise even for a video signal having an image portion whose gradation change is gradual such as gradient. Objective.

  Another object of the present invention is to provide a block noise reduction device that can reduce block noise without impairing the high frequency component of the video signal.

  Furthermore, another object of the present invention is to provide a block noise reduction apparatus capable of stopping image quality degradation by stopping the block noise reduction processing for a video signal in which the steps at both ends of the rectangular block are not block noise. It is in.

In order to achieve the above object, a block noise reduction apparatus according to a first aspect of the present invention reduces block noise from a video signal obtained by decoding a digital video signal compressed and encoded in units of rectangular blocks in units of rectangular blocks. In the block noise reduction apparatus, the calculation means for calculating the first and second boundary step values at both ends of the rectangular block for each rectangular block of the video signal before the block noise reduction, and the calculation means calculated by the calculation means. Based on the values of the first and second boundary steps, the values in the first boundary step and the values in the second boundary step are connected according to an arbitrary function set in advance as the value of each pixel in the rectangular block. A correction means for calculating such a value, adding the value to the video signal before the block noise reduction, and outputting it, and first and second ends of the rectangular block of the video signal before the block noise reduction. If the second boundary step value is compared with a preset threshold value to determine whether block noise has occurred at both ends of the rectangular block, and if it is determined that block noise has occurred Determining means for outputting a switch control signal having a second predetermined value different from the first predetermined value when it is determined that the first predetermined value has not occurred;
If the value of the switch control signal from the determination means of the first predetermined value, select the video signal output from the correction means, in the case of the second predetermined value, predetermined in block noise reduction before the video signal And a selecting means for selecting and outputting a video signal obtained by filtering the characteristic.
Here, the determination means obtains a result of comparing the first and second boundary step values at both ends of the rectangular block of the video signal before the block noise reduction with a preset threshold value. A switch control signal having a first predetermined value may be output when both of the second boundary step values are equal to or less than the threshold value, and a second predetermined value may be output otherwise.

In the present invention, whether or not block noise is generated at both ends of the rectangular block according to the comparison result between the first and second boundary step values at both ends of the rectangular block of the video signal before the block noise reduction and the threshold value. If it is determined that block noise has occurred, the video signal output from the correction means is selected. If it is determined that block noise has not occurred , the video signal or block before block noise reduction is selected. the video signal obtained by filtering the predetermined characteristic in the noise reduction before the video signal can be output by selecting.

  According to the present invention, based on the values of the first and second boundary steps at both ends of the rectangular block, the value in the first boundary step and the value in the second boundary step as the value of each pixel in the rectangular block. To obtain a video signal with reduced block noise by calculating a value that is connected to the value of the video signal according to an arbitrary function set in advance and adding the value to the video signal before the block noise reduction. For this reason, block noise generated in an image portion where gradation change is gradual such as gradient can be sufficiently reduced.

  Further, according to the present invention, since the block noise is reduced by correcting the DC component of the rectangular block without using a filter, the high frequency component of the image when using the filter does not deteriorate. Block noise can be reduced.

  Further, according to the present invention, when it is determined that the block noise is generated at both ends of the rectangular block by comparing the first and second boundary step values with the threshold value, the block noise reduction processing is performed. When a video signal is output and it is determined that block noise has not occurred, select the video signal before block noise reduction or the video signal obtained by filtering the video signal before block noise reduction with the specified characteristics. Therefore, it is possible to avoid significant image quality degradation caused by performing block noise reduction processing on a video signal in which no block noise has occurred.

  Next, the best mode for carrying out the present invention will be described. FIG. 1 shows a block diagram of a first embodiment of a block noise reduction apparatus according to the present invention. As shown in the figure, in the present embodiment, a video signal before block noise reduction obtained by decoding a digital video signal compressed and encoded in units of rectangular blocks in units of rectangular blocks is supplied. It comprises a correction circuit 11 and a block noise level difference calculation circuit 12.

  The block noise level calculation circuit 12 calculates the block noise edge level values at both ends of the rectangular block of the input video signal before block noise reduction, and outputs the block noise edge level value to the correction circuit 11. The correction circuit 11 performs a correction operation based on a predetermined correction formula for the input video signal before block noise reduction based on the block noise edge step value input from the block noise step calculation circuit 12. To output a video signal from which block noise has been reduced or eliminated.

  Next, the operation of the present embodiment will be described in detail with reference to FIG. A block boundary line of interest in the horizontal direction or the vertical direction of the video signal before block noise reduction supplied to the correction circuit 11 and the block noise level calculation circuit 12 is indicated by a dotted line I in FIG. And Further, it is assumed that the block boundary line of adjacent rectangular blocks is n pixels after the point of interest as indicated by a dotted line II in FIG. In the present invention, it is assumed that the location of the boundary line of block noise is known in advance from information such as a digital tuner.

  In this case, the block boundary I of interest is between the last pixel 0 of the first rectangular block and the first pixel 1 of the second rectangular block in FIG. The level difference is A. Further, the adjacent block boundary line II is between the last pixel n of the second rectangular block having n pixels in the horizontal direction or the vertical direction and the first pixel n + 1 of the adjacent third rectangular block. There is a step B which is the difference between the pixel levels.

  The block noise level calculation circuit 12 calculates an intermediate value “a” of the level difference A between the rectangular blocks at the block boundary line I of interest and an intermediate value “b” of the level difference B between the rectangular blocks at the adjacent block boundary line II by the following equations. calculate.

a = (P0−P1) / 2 (1)
b = (Pn + 1−Pn) / 2 (2)
In the equations (1) and (2), P0, P1, Pn + 1, and Pn indicate the pixel levels of the pixels 0, 1, n, and n + 1, respectively.

  The correction circuit 11 is supplied with the above-described intermediate values a and b calculated by the block noise step calculation circuit 12 (these are the intermediate values of the block noise edge steps at both ends of the second rectangular block). Based on this, the correction calculation is performed on the input video signal by the following equation.

    So = Si + a + (b−a) × (2 × k + 1) / (2 × n) (3) where, in equation (3), So is the output video signal level of the correction circuit 11, and Si is the input video of the correction circuit 11. The signal level, k is a variable whose value changes corresponding to the pixel position in the rectangular block, and n is the number of pixels in the horizontal or vertical direction of the rectangular block.

  Here, the value of the variable k is 0 for the pixel 1, 1 for the pixel 2, and the value of the variable k monotonously increases by 1 as the pixel position advances to the final pixel position in the rectangular block. n takes a value of n-1. The correction circuit 11 performs the correction calculation of the above equation (3) according to the pixel position of the second rectangular block, so that the correction circuit 11 receives the intermediate value a as shown in FIG. A video signal having a level So in which the pixel levels of the pixels are aligned on a straight line connecting the intermediate value b and the intermediate value b is output.

  Similarly, for the other rectangular blocks, the correction circuit 11 calculates a value obtained by connecting the intermediate values of the respective steps at both ends of the rectangular block with a straight line, and outputs the video signal obtained by adding the value to the input video signal. As a result, as can be seen from FIG. 2B, the correction circuit 11 outputs a video signal with sufficiently reduced block noise, even for a video signal having an image portion whose gradation change is gradual such as gradient. Is done.

  In this embodiment, since the DC component of the video signal of the rectangular block is corrected without using the filter, the high frequency component of the image generated when the filter is used is not impaired. Noise can be reduced or eliminated.

  Next, a second embodiment of the present invention will be described. FIG. 3 shows a block diagram of a second embodiment of a block noise reduction apparatus according to the present invention. In the figure, the same components as in FIG. Compared with the first embodiment, the second embodiment shown in FIG. 3 is characterized in that a block noise determination circuit 13 and a switch circuit 14 are added. The switch circuit 14 is switching-controlled by a switch control signal from the block noise determination circuit 13.

  In FIG. 3, the video signal before block noise reduction obtained by block decoding the block-coded digital video signal is supplied to the correction circuit 11 and the block noise level calculation circuit 12, respectively, and the block noise The determination circuit 13 and the two-input switch circuit 14 are respectively supplied to one input terminal 14b. The corrected video signal output from the correction circuit 11 is supplied to the other input terminal 14 a of the switch circuit 14.

  The block noise determination circuit 13 obtains block noise edge steps at both ends of the rectangular block for each rectangular block of the supplied video signal before the block noise reduction, compares them with a threshold value, and according to the comparison result Outputs a logical switch control signal. Now, assuming that the block noise level difference at one end of the rectangular block is A, the block noise level difference at the other end after n pixels from the one end is B, and the above threshold is Th, the block noise determination circuit 13 One of the following four comparison results is obtained.

(1) A ≦ Th, B ≦ Th
(2) A> Th, B ≦ Th
(3) A ≦ Th, B> Th
(4) A> Th, B> Th
Here, there is a video signal in which the steps at both ends of the rectangular block are not block noise. In such a video signal, the high frequency component is reduced when the block noise occurs, whereas the high frequency component with a large level difference remains. Is likely to be the original signal. Therefore, in the present embodiment, the block noise is determined and corrected only when the steps at both ends of the rectangular block are both lower than the threshold value.

  Therefore, the block noise determination circuit 13 performs the switch circuit 14 only when the comparison result (1) in which both of the two block noise steps A and B are equal to or less than the threshold Th among the above four comparison results is obtained. When a switch control signal having a first logical value (for example, H) for selecting the input signal of the terminal 14a is output and other comparison results (2), (3) or (4) are obtained, the switch circuit 14 outputs a switch control signal of a second logic value (for example, L) that selects the input signal of the terminal 14b.

  As a result, from the common terminal 14c of the switch circuit 14, for the rectangular block in which both the two block noise steps A and B are both equal to or less than the threshold value Th, the video signal subjected to the block noise reduction processing by the correction circuit 11 is extracted. For a rectangular block in which one or both of the two block steps A and B or both are larger than the threshold Th, it is determined that the block noise is not block noise, and the video signal (original signal) before block noise reduction is extracted as it is. It is.

  As described above, according to the present embodiment, the block noise reduction process is not performed on the video signal determined not to be the block noise, so that the block noise reduction is performed on the video signal that is not the block noise. Significant image quality degradation caused by the processing can be avoided.

  Next, a third embodiment of the present invention will be described. FIG. 4 shows a block diagram of a third embodiment of a block noise reduction apparatus according to the present invention. In the figure, the same components as those in FIG. Compared with the second embodiment, the third embodiment shown in FIG. 4 is characterized in that a filter 15 is added and the operation of the switch circuit 16 is slightly different from that of the second embodiment. The switch circuit 16 is switching-controlled by a switch control signal from the block noise determination circuit 13.

  In FIG. 4, the video signal before block noise reduction obtained by block decoding the block-coded digital video signal is supplied to the correction circuit 11, the block noise level calculation circuit 12, and the block noise determination circuit 13, respectively. At the same time, the signal is supplied to one input terminal 16 b of the two-input switch circuit 16 through the filter 15. The corrected video signal output from the correction circuit 11 is supplied to the other input terminal 16 a of the switch circuit 16.

  The filter 15 is a low-pass filter for reducing noise including block noise. For example, as shown in FIG. 5, delay circuits 151 and 152 for delaying the input video signal by one pixel and the input video signal are provided. A multiplier 153 that multiplies the coefficient by ¼, a center tap multiplier 154 that multiplies the video signal output from the delay circuit 151 by one pixel by a coefficient ½, and a delay circuit 152. And a multiplier 155 that multiplies the video signal delayed by a total of two pixels by a coefficient ¼, and an adder 156 that adds the output signals of the multipliers 153, 154, and 155. .

  Thereby, in the present embodiment, for the rectangular block in which both of the two block noise steps A and B at both ends of the rectangular block are equal to or less than the threshold Th, the video signal subjected to the block noise reduction processing by the correction circuit 11 is received. It is taken out from the common terminal 16 c of the switch circuit 16. On the other hand, for a rectangular block in which one or both of the two block noise steps A and B or both are larger than the threshold value Th, it is determined that the block noise is not block noise, and the image in which noise is reduced by the filter processing by the filter 15 A signal is taken out from the common terminal 16 c of the switch circuit 16.

  As described above, according to the present embodiment, for the video signal determined not to be block noise, the noise reduction processing is performed by the filter 15 and the block noise reduction processing is not performed. It is possible to avoid significant image quality degradation caused by performing block noise reduction processing on a non-video signal.

Next, a fourth embodiment of the present invention will be described. FIG. 6 shows a block diagram of a fourth embodiment of a block noise reduction apparatus according to the present invention. In the figure, the same components as those in FIG. 4 are denoted by the same reference numerals, and the description thereof is omitted. The fourth embodiment shown in FIG. 6 is slightly different from the third embodiment in the configuration and operation of the block noise determination circuit 17 and the configuration and operation of the switch circuit 18 as compared with the third embodiment. There is a feature in the point.

  In FIG. 6, the 3-input 1-output type switch circuit 18 is supplied with the output signal of the correction circuit 11, the video signal before block noise reduction, and the output signal of the filter 15 at three input terminals 18a, 18b, and 18c, respectively. One input signal is output to the outside via the common terminal 18d by the switch control signal generated by the block noise determination circuit 17.

  The block noise determination circuit 17 obtains block noise edge steps at both ends of the rectangular block for each rectangular block of the supplied video signal before block noise reduction, and the two threshold values Th1 and Th2 (where Th1> Th2). ) And a switch control signal having a logical value corresponding to the comparison result is output.

  Here, assuming that the block noise level difference at one end of the rectangular block is A and the block noise level difference at the other end after n pixels from the one end is B, the block noise determination circuit 17 performs the above block noise level difference A and When both of B are lower than the threshold value Th2 having a smaller value, it is determined as block noise, and the switch circuit 18 is supplied to the terminal 18a and the video signal subjected to the block noise reduction processing by the correction circuit 11 is selected. Switching control is performed so that

  Further, when both the block noise steps A and B are higher than the larger threshold value Th1, the block noise determination circuit 17 determines that the block noise is not a block noise but a video signal, and connects the switch circuit 18 to the terminal. Switching control is performed so that the video signal before block noise reduction supplied to 18b is selected. Further, the block noise determination circuit 17 obscures the block noise or the video signal when both or one of the block noise steps A and B is a value between the threshold values Th1 and Th2. Switching control is performed so that the switch circuit 18 selects a video signal that has been subjected to noise reduction processing by filtering without performing correction for reduction.

  Thus, according to the present embodiment, since the processing by the correction circuit 11 for reducing block noise is not performed on the video signal that is not block noise, the block noise reduction processing is performed on the video signal that is not block noise. In addition, it is possible to avoid significant image quality degradation caused by performing image processing, and to output a video signal that has been subjected to noise reduction processing by the filter 15 for a block noise or video signal that is ambiguous. A video signal with reduced block noise can be output.

  The present invention is not limited to the above embodiment. For example, in FIG. 4 and FIG. 6, the filter 15 has been described as having the same characteristics, but differs depending on each embodiment. The one with the optimum characteristics can be used. The threshold values in the block noise determination circuits 13 and 17 can be changed and set as necessary.

  Further, in the above embodiment, the correction circuit 11 calculates a value connecting the intermediate values of the respective steps at both ends of the rectangular block with a straight line, and adds the value to the input video signal. What is necessary is just to calculate the value which connects the value in each level | step difference of the both ends of a rectangular block according to the preset arbitrary function, and just add the value to an input video signal. As said arbitrary function, the straight line of embodiment In addition, the value in each step at both ends of the rectangular block is not limited to the intermediate value of the step in the embodiment.

It is a block diagram of a 1st embodiment of the present invention. It is explanatory drawing of an example of the block noise reduction operation | movement in this invention. It is a block diagram of the 2nd Embodiment of this invention. It is a block diagram of the 3rd Embodiment of this invention. It is a circuit diagram of an example of the filter in FIG. It is a block diagram of the 4th Embodiment of this invention. It is explanatory drawing of the subject by the conventional block noise reduction.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 Correction circuit 12 Block noise level | step difference calculation circuit 13, 17 Block noise determination circuit 14, 16, 18 Switch circuit 15 Filter

Claims (2)

In a block noise reduction device for reducing block noise from a video signal obtained by decoding a digital video signal compressed and encoded in units of rectangular blocks in units of rectangular blocks,
Calculating means for calculating values of first and second boundary steps at both ends of the rectangular block for each rectangular block of the video signal before block noise reduction;
Based on the values of the first and second boundary steps calculated by the calculation means, the value in the first boundary step and the value in the second boundary step as the value of each pixel in the rectangular block. A correction unit that calculates a value that connects the value of the value according to an arbitrary function set in advance, adds the value to the video signal before the block noise reduction, and
Block noise occurs at both ends of the rectangular block by comparing the first and second boundary step values at both ends of the rectangular block of the video signal before the block noise reduction with a preset threshold value. If it is determined that the block noise is generated, the first predetermined value is determined. If it is determined that the block noise is not generated, the second predetermined value is different from the first predetermined value. Determining means for outputting a switch control signal;
Wherein when the value of the switch control signal from the determining means of said first predetermined value, said selected video signal output from the correction means, the case of the second predetermined value, before Symbol block noise reduction A block noise reduction apparatus comprising: selection means for selecting and outputting a video signal obtained by filtering the previous video signal with a predetermined characteristic.   The determination means compares the first and second boundary step values at both ends of the rectangular block of the video signal before the block noise reduction with a preset threshold value. The switch control signal of the first predetermined value is output when both second boundary step values are equal to or less than the threshold value, and the second predetermined value is output otherwise. The block noise reduction device according to 1.

Images