JPH06282946A - Picture signal reproducing device - Google Patents

Abstract

PURPOSE:To obtain a picture signal reproducing device capable of a data form- correction even when a motion error detection and the omission of an adjacent block are present. CONSTITUTION:Form-correction data 107, 111, 128 being respectively presedent by one frame presedent by one field of and adjacent to a present signal 106 are made and the center value is made to be a form-correction completion data 123 by being selected by a center value selecting circuit 122 and the data is fed-back in accordance with a flag 102 indicating a data omission block via a switch 124. Further, when the form-correction data 128 is of an omission block, the data is replaced with a fixed value 130. Thus, the purpose is attained with a relatively simple circuit, the data form-correction minimizing the degradation of pictures is performed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used in an image signal recording / reproducing apparatus such as a digital VTR, especially when an error of image data in which an image signal is block-encoded appears as a block-shaped loss on the screen. The present invention relates to an image signal reproducing device for modifying data.

[0002]

2. Description of the Related Art FIG. 4 shows an outline of a signal processing section of a conventional image signal recording / reproducing apparatus.

In FIG. 4, during recording, the input image signal 20 is converted into a digital signal by an A / D converter 21,
After the block coding unit 22 performs compression or the like on a block-by-block basis composed of a plurality of pixels, an error correction code adding circuit 23 adds an error correction code and records it on a recording medium 24 such as a tape or a disk.

At the time of reproduction, after the error on the recording medium 24 is corrected by the error correction unit 25, block decoding is performed by the block decoding unit 26. If the error correction cannot be performed, the position is changed. Uncorrectable flag 29
Is sent together with the data, is interpolated by the data adjusting unit 27 using peripheral pixels, etc., and is output as a reproduced image signal 30 through the D / A converter 28.

FIG. 5A is a block diagram showing a first conventional example of the data modifying unit 27. In the figure, when the reproduced data 301 has a missing block, the switch 302 is controlled by the flag 29 and the frame is changed. It is replaced with the data one frame before that has passed through the memory 303 and is output as data 304.

FIG. 6 is a side view of the scanning line, and the state of data modification can be visually observed. FIG. 6A shows the first
In the related art, the pixels 401 in the missing blocks indicated by black circles are replaced with the pixels 402 one frame before. For the sake of explanation, the size of one block is 4 pixels (lines) here.

FIG. 5B is a block diagram showing a second conventional example of the data modifying section 27. In the figure, when the reproduced data 311 has a missing block, the switch 312 is controlled by the flag 29 to modify the modified data. To replace
This is the same as the conventional example, but this time, two types of modification data are prepared. First, the interframe correction circuit 314 obtains a signal one frame before, as in the first conventional example. Next, the intrafield correction circuit 313 creates interpolation data from peripheral pixels in the same field as the missing block. These two
In order to select appropriate data from the types of modification data, the motion detection circuit 315 determines whether the image at the missing block position has a motion (hereinafter a moving image) or not (hereinafter a still image). Based on this result, the switch 316 is controlled to select the in-field modification data for a moving image and the inter-frame modification data for a still image, and the switch 312 is selected.
And output as data 318.

Since the data of the missing block is not literally correct as a method of motion detection, there is no choice but to estimate it by using a frame difference between peripheral blocks adjacent to the missing block or peripheral pixels. Further, the frame memory used in the motion detection circuit 315 can share the frame memory inside the interframe correction circuit 314.

In FIG. 6B, the missing pixel 411 with a black circle is shown.
Shows a case in which is replaced by the pixel 412 one frame before, and a case in which the pixel is calculated from the adjacent upper and lower pixels 413 and 414 in the same field and interpolated.

[0010]

In the first prior art example described above, the processing is relatively simple and a good retouching result can be obtained in the case of a still image, but an unnatural image in the case of a moving image. There is a problem. Further, although the second conventional example improves the drawbacks of the first conventional example, there are problems that the circuit scale of the motion detection circuit 315 becomes large and image quality deterioration due to erroneous detection. This erroneous detection results in deterioration of detection accuracy due to the fact that the motion of the missing block itself cannot be detected and is estimated from the amount of motion of the surroundings, and when the peripheral blocks of the target block are also missing, the amount of motion The main reason is that calculation becomes impossible.

Further, in both the first and second conventional examples, since data modification is performed in block units, a large amount of memory is required when modification is performed after the data is in the raster scan order. . In particular, when a block extends over a field and is collectively modified, the circuit becomes a very large scale.

The present invention has been made in order to solve the above problems, and provides an image signal reproducing apparatus capable of performing good data correction even if an erroneous motion is detected or a neighboring block is missing. Is intended.

[0013]

In the present invention, a flag generating means for generating a flag indicating the position of a block to be retouched, and data of one frame before the retouching target block, one field before, and an adjacent block are used. Correction data generating means for generating a plurality of correction data in pixel units, a median value selecting means for selecting a median value or a vicinity of the median value of the plurality of modification data, and input data according to a flag from the flag generating means. Switch means for adding data or input data obtained by replacing missing data in the median value selection means with output data from the median value selection means, and whether each of the modification data is valid according to a flag from the flag generation means. If the judgment means for judging whether or not the judgment by the judgment means is invalid, the retouched data is judged by the median value selection means. The invalid data removing means for preventing from being-option is provided.

[0014]

Operation: At least three types of retouching data are prepared for each pixel in the block where the data is missing, and the median value is selected from the retouching data to be retouched data. In addition, the modified data is fed back by the switch means so that it can be used for the subsequent block modification, and the modified data created from the data missing block is excluded from the selection of the median value.

[0015]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows an embodiment of the image data adjusting section of the image signal reproducing apparatus according to the present invention. In FIG. 1, 10
Reference numeral 1 is reproduced data after error correction, and data that could not be error corrected is included as a missing block. 102
Is a flag indicating the position of the missing data block, and FIG.
This is the same as the uncorrectable flag 29 from the error correction circuit 25. In the present embodiment, the size of the data missing block will be described as 4 lines × n (n is not particularly defined).

Delay memories 103, 104 and 105 have delay amounts of 1H, 258H and 262H, respectively. Here, 1H indicates a delay of one line.

In this embodiment, these delay memories 10 are
Three types of modification data are created from the taps 3, 104, and 105, and one of them is selected as final modification data.

First, assuming that the output signal 106 of the delay memory 103 at the 4th H represents the present signal, a signal of one frame (525H) before is obtained as the output signal 107 of the delay memory 105. Data. In FIG. 2, assuming that the pixel 421 is the current signal 106, the pixel 4
22 corresponds to the signal 107 one frame before.

Next, the output signal 108 of the delay memory 104
Since signals 262H before and 263H before, that is, a signal of one field before are obtained as the output signal 109 of the delay memory 103 and the next, respectively, these signals are averaged by the adder 110, and the signal 111 of the addition result is obtained. This is the second modification data. In FIG. 2, assuming that the pixel 421 is the current signal 106, the pixels 423 and 424 are 262H before and 263H respectively.
It becomes the previous signal.

The third modification data is created from the peripheral pixels of another block in the same field of the missing block. Here, retouching data is created from two pixels vertically adjacent to the missing block by a linear interpolation method.

In FIG. 2, the pixel 421 is connected to the current signal 106.
Then, two pixels vertically adjacent to the block, that is, a pixel 425 on two lines and a pixel 426 on three lines below are used. The pixels 425 and 426 are the output signals 11 of the delay memory 103 at the first and sixth stages in FIG.
It corresponds to 2,113. By passing these through a weighting adder 114 that is proportional to the distance between pixels, modification data 116 by linear interpolation is obtained. In this example, the current signal 106
The pixel 421 of No. 2 is in the second line from the top in the block as shown in FIG. 2, but in the case of the first, third, and fourth lines, the retouching data 115, 117, and 118 are similarly obtained. The selector 119 uses the four modification data 115-115.
From 118, one of them is selected by an in-block line number 120 (input from a counter (not shown)) indicating which line the current signal 106 is in the block, and third modification data 121 is obtained.

As described above, the three types of modification data 107, 11
1 and 121 are obtained, but when the pixel data that is the basis of these calculations is missing, that is, in FIG. 2, pixels 422, 423, 424, 425, 426, etc. are missing. Correct data cannot be obtained. Therefore, the following measures are taken.

First and second modification data 107, 11
With respect to 1, since the signals one frame before and one field before are used as the original data of the adjustment data, even if these original data are missing data, they must have already been adjusted. is there. Therefore, when the current signal 106 is written in the next delay memory 103, if it is missing data, it may be replaced with the modification data already obtained. The modification data is the median selection circuit 1
It is available at the output 123 of the switch 22.
Give feedback through 4.

The switching of the switch 124 is controlled by the signal 126 obtained by delaying the flag 102 by 4 lines in the 4H memory 125 in order to match the timing with the data. As a result, the output signal 129 of the switch 124 becomes the modified data, and the pixels 422, 423, 424, 4 of FIG.
In 25, correct data or modified data is always obtained.

Next, regarding the third modification data, the pixel 426 of FIG. 2 as the original data is future data, and the above method cannot be used.
If 26 is missing data, the first or second modification data is used as the modification data. In order to realize this, first, a switch 127 used as invalid data removing means is provided to input the third modification data 121, and the modification data is controlled to a predetermined value when the pixel 426 is missing by controlling with the flag 102. Is replaced with a fixed value 130. This fixed value is a value for automatically removing the third modification data when the median value is selected later, and the maximum value or the minimum value that the data can take is used.

The output signal 128 of the switch 127 obtained as a result is newly used as the third modification data.

As described above, the three types of modification data 107, 11
Since 1 and 128 are obtained, the median value selection circuit 12
2 is input to select the final modification data 123 and is input to the switch 124.

FIG. 3 shows a block diagram of the median selection circuit 122. In FIG. 3, the three types of input data 107, 111, 128 are first compared in size by the three comparators 51, and the comparison result is determined by the decoder 52.
The selector 53 selects the median value and outputs the modification data 123.

Referring again to FIG. 1, since the modified data has already been obtained as the signal 129 from the switch 124, this is output from the output terminal 131 to be the final output of this embodiment.

In the above description, the third modification data 1
When 28 is invalid, it is replaced with a fixed value 130 so as not to be selected and input to the median selection circuit 122. At this time, however, the median selection circuit 122 is not used, and for example, the first
It is also possible to use the modification data 107 of 1. In addition, in this embodiment, three types of data 10
7, 111, and 128 are obtained from the pixels between the frames, between the fields, and the upper and lower pixels in the field, but the invention is not limited to these, and for example, the left and right pixels of the block in the field are used. It is also possible to use three or more types of modification data.

[0032]

As described above, according to the image signal reproducing apparatus of the present invention for selecting the median value from a plurality of modification data for each pixel, the image quality deterioration due to the erroneous motion detection can be achieved with a relatively simple circuit. Therefore, even if the data of the adjacent block is missing, good retouching is possible. Further, even if the retouching data which is not appropriate for the image quality is selected, the median value is used, so that the deterioration of the image quality can be minimized.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing an embodiment of the present invention.

FIG. 2 is a configuration diagram showing pixels of each line for explaining the operation of the embodiment.

FIG. 3 is a block diagram showing an embodiment of a median selection circuit.

FIG. 4 is a block diagram of a conventional image signal recording / reproducing apparatus.

FIG. 5 is a block diagram showing an example of a conventional data modifying unit.

FIG. 6 is a configuration diagram showing a conventional data modification method.

[Explanation of symbols]

 103, 104, 105 Delay memory 107, 111, 128 Modification data 110, 114 Adder 122 Median value selection circuit 124, 127 switch

Claims (1)

[Claims] 1. An image signal reproducing apparatus for correcting, in pixel units, data in a block in which data is missing when decoding image data encoded in block units composed of a plurality of pixels. A flag generating means for generating a flag indicating the position of, and a retouching data generating means for generating a plurality of retouching data pixel by pixel using the data of one frame before the retouching target block, one field before, and the adjacent block, Median value selecting means for selecting the median value or the vicinity of the median value of the plurality of modification data, and data obtained by replacing the missing data in the input data with the output data of the median value selecting means according to the flag from the flag generating means. Alternatively, a switch means for adding input data to the modification data generating means, and each modification described above according to a flag from the flag generating means. A determination means for determining whether or not the data is valid, and an invalid data removal means for preventing the modification data from being selected by the median value selection means when the determination by the determination means is invalid are provided. An image signal reproducing device characterized by.