JPH0846961A - Method for encoding and decoding picture - Google Patents

Abstract

PURPOSE:To eliminate boundary line conspicuousness caused by the difference in decoding picture quality at both boundary sides by respectively giving weighting averages to the overlapping areas of small pictures so that connection is smoothed in a TV picture obtained by means of compositing the decoded small pictures. CONSTITUTION:At a transmission side, an HDTV signal 202 from an input terminal 201 is divided into the two small pictures 205 and 206 through the use of a picture dividing part 203 and division is executed so as to contain the two division pictures in the overlapping area concerning the area which is set by an overlapping area setting part 204. The respective small pictures are inputted to information compressing parts 207 and 208 to execute an encoding processing and encoded outputs 209 and 210 are inputted to a multiplexing part 211 and collected into one encoding data 212. At a reception side, received encoded data 215 is separated into encoding data 217 and 218 as against the small pictures which are separated through the use of a multiplex separating part 216 and the decoding processing is respectively executed in information processing part 219 and 220. Then, data of overlapping decoded pictures 225 and 226 is transmitted to a weighting average part 229 and outputted as the final HDTV signal 232.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a high-efficiency coding / decoding method for dividing a signal having a high sampling rate, such as an HDTV image, into a plurality of small images for parallel processing. HDTV obtained by combining images
The present invention relates to an image coding / decoding method that avoids detection of a boundary line between small images on a decoded image such as an image, and as a result can obtain a decoded image with good image quality.

[0002]

2. Description of the Related Art When a signal sampled at a high frequency such as an HDTV image is encoded with high efficiency, a parallel processing method is generally used to operate hardware in real time.

Consider a case where an HDTV signal is spatially divided into n to form a small image, and the encoding processing is performed in parallel. On the encoding side (transmission side), one screen of an HDTV signal is first divided into n small images that do not overlap each other, and each small image is encoded in parallel. The coded data of n small images that have been coded are subjected to decoding processing on the decoding side (reception side), and then placed at corresponding positions on the HDTV screen,
The combined signal becomes the decoded HDTV signal.

FIG. 6 shows an example of a conventional screen division type image coding / decoding method when n = 2. The block configuration of parallel encoding in this case is as shown in FIG. In FIG. 7, 701 is an input terminal, 702 is an input HDTV signal, 703 is an image dividing unit, 704 and 705 are divided standard TV size small images, 706 and 707 are information compressing units,
708 and 709 are encoded outputs, 710 is a multiplexing unit, 71
1 is multiplexed encoded data, 712 is an output terminal, 7
Reference numeral 13 is a transmission line, 714 is a receiving terminal, 715 is received encoded data, 716 is a demultiplexing unit, 717 and 718 are encoded data for separated small images of standard TV size, and 719 and 720 are information decoding units. , 721 and 722 are the decoded small images, 723 is the image synthesizing unit, and 724 is the decoded H.
DTV signal, 725 represents an output terminal.

First, on the transmitting side, the HDTV signal 702 input from the input terminal 701 is divided into two small images 704 and 705 by an image dividing unit 703. Each of the divided small images 704 and 705 has an information compression unit 7
It is input to 06, 707 and subjected to encoding processing. The encoded outputs 708 and 709 are input to the multiplexing unit 710, and 1
After being combined into one encoded data 711, the output terminal 7
It is sent to the transmission line 713 via 12.

On the other hand, on the receiving side, the coded data 715 received from the transmission line 713 at the receiving terminal 714 is separated into the coded data 717 and 718 corresponding to the two small images by the demultiplexing unit 716. The separated encoded data 717 and 718 are respectively separated by the information decoding unit 71.
The data is input to 9,720 and subjected to a decoding process. The decoded small images 721 and 722 are input to the image synthesis unit 723, restored to the size of the HDTV signal, and output to the output terminal 725 as the decoded HDTV signal 724.

[0007]

In the method as described above, the coding and decoding of the divided small images are processed independently. Therefore, when the image characteristics such as the presence or absence of fine parts, the size and complexity of movements, etc. differ between small images, the encoding parameters of the fineness of quantization are different for each small image. As a result, the way in which coding noise occurs varies from small image to small image.

Further, in the encoding process, it is desirable from the viewpoint of image quality that the encoding is performed using substantially the same quantization characteristic in a continuous area and an area having similar image characteristics. . However, since the coding and decoding of the divided small images are independent of each other, the continuity of the coding process in the boundary region between the small images, which is an originally continuous region, is lost, resulting in a completely different quantization. There is a possibility that it will be encoded in the property.

FIG. 8 is a diagram showing how noise is generated at a boundary portion of a decoded image synthesized by the conventional method, and a straight line connecting A and B on the screen of the encoded and decoded HDTV signal. It is an example of the decoded value level of the upper pixel. Since the small image to which A belongs and the small image to which B belongs are different from each other, they are encoded independently, and as a result, the small image to which B belongs is considerably coarser than the small image to which A belongs. If it is used, the noise level of the decoded image is different between the left and right of the boundary point C of both small images as shown in FIG.

From the above, according to the conventional method of simply inserting the encoded and decoded small image at the corresponding position on the HDTV screen on the decoding side, the decoding HDTV
There was a problem that the boundaries between the small images were detected on the screen, resulting in a decoded image with a feeling of strangeness.

[0011]

The present invention has been made to solve this problem, and the image coding / decoding method of the present invention, on the coding side, produces a plurality of small images. When dividing, the boundary areas are divided so that they overlap each other, and encoding and decoding are performed for each divided small image. On the other hand, in decoding, when synthesizing encoded small images, for a region that does not overlap, the decoded value of the small image to which the region belongs is the final reproduction value, but for the overlapping region Performs weighted averaging on the decoded values of each small image including that area to obtain the final reproduction value.

FIG. 1 shows an example of a screen division type image coding / decoding method according to the present invention in the case of two divisions. Figure 2
Shows an example of a block configuration for implementing this example. Figure 2
, 201 is an input terminal, 202 is an input HDTV signal, 203 is an image division unit, 204 is an overlap region setting unit, 2
Reference numerals 05 and 206 denote divided small images, 207 and 208 information compression units, 209 and 210 coded outputs, 211 multiplexing units, 212 multiplexed coded data, 213 output terminals, and 214 reception. A terminal, 215 is received encoded data, 216 is a demultiplexing unit, 217 and 218 are encoded data for the separated small images, 219 and 220 are information decoding units, and 221 and 222 are codes for the decoded small images. Data, 223, 224 are duplication determination sections, 225,
226 is an overlap decoded image for the small image, 227, 228
Is a non-overlapping area of each small image, 229 is a weighted averaging portion, 230 is a weighted coefficient processed image of overlapping decoded images, 231 is an image synthesizing portion, 232 is the finally decoded HDTV signal, and 233 is Represents an output terminal.

On the transmitting side, the HDTV signal 202 input from the input terminal 201 is divided into two small images 205 and 206 by the image dividing unit 203. At this time, the area set by the overlapping area setting unit 204 is divided so as to be included in the two divided images. The divided small images are input to the information compression units 207 and 208, respectively, and subjected to encoding processing. Encoded outputs 209, 2
10 is input to the multiplexing unit 211, put together into one encoded data 212, and then output to the output terminal 213.

On the receiving side, from the transmission line to the receiving terminal 2
The encoded data 215 received by the demultiplexer 2
The encoded data 21 corresponding to two small images in 16
Separated into 7,218. The encoded data 217 and 218 that have been separated are information decoding units 219 and 220, respectively.
And is decrypted. Areas 227 and 228 of the decoded small images which are not overlapping portions are input to the image synthesizing unit 231 as they are. In addition to this, in the overlap determination units 223 and 224, the encoded data of the overlap decoded images 225 and 226 that are determined to be the overlap portions are sent to the weighted averaging unit 229. Weighted averaging part 2
In 29, the weighted average value of the decoded values of both small images in the area corresponding to the overlapping portion is calculated, and the weighted coefficient processed image 230 of the overlapping decoded image is the area 22 which is not the above-mentioned overlapping portion.
7, 228 are input to the image synthesizing unit 231, and HD
The HDTV signal 232 of the final decoded image is obtained at the output terminal 233 by being fitted into the corresponding position of the TV signal.

[0015]

According to the method as described above, in an HDTV image formed by synthesizing decoded small images, it is possible to apply a weighted average to each overlapping region of the small images and smoothly connect them. Therefore, it is possible to make the boundary line, which is caused by the difference in the properties of the decoded image on both sides of the boundary, inconspicuous.

As a result, it is possible to effectively make it difficult to detect the boundary line of each small image in the HDTV image,
It is possible to improve the image quality of the decoded image.

[0017]

Examples of the present invention will be described below. In the present embodiment, consider a case where an HDTV signal is divided into two small images for processing and discrete cosine transform is used as an information compression method. As shown in FIG. 3, on the transmission side, the HDTV image F (i, j) (i = 1,2, ..., 2M, j = 1,2, ..., N) is horizontally 2x (x> 0). The two small images f ₁ (i, j) and f ₂ (i, j) are overlapped with each other.

At this time, f ₁ (i, j) = F (i, j) (1≤i≤M + x) f ₂ (i, j) = F (i, j) (M-x + 1≤i≤2M) Yes, the part of M−x + 1 ≦ i ≦ M + x is an overlapping part. The encoded and decoded image of each small image is f ₁ ′
(i, j), f ₂ ′ (i, j) Decoded HDTV image F ′
(i, j) is calculated as follows. F ′ (i, j) = f ₁ ′ (i, j) (1 ≦ i ≦ M−x) F ′ (i, j) = w ₁ (i, j) f ₁ ′ (i, j) + w ₂ (i, j) f ₂ ′ (i, j) (M−x + 1 ≦ i ≦ M + x) F ′ (i, j) = f ₂ ′ (i, j) (M + x + 1 ≦ i ≦ 2M) where w ₁ ( If i, j) = w ₂ (i, j) = 1/2, it means that the average value of both decoded small images is taken at any position in the overlapping area.

Further, when the weighted average is applied to the decoded values of the overlapping small images, the weighting coefficient can be determined for each pixel by the pixel position in the overlapping area. When w ₁ and w ₂ are set as w ₁ (i, j) = − i / (2x + 1) + (M + x + 1) / (2x + 1) w ₂ (i, j) = 1−w ₁ (i, j) Means that the weight becomes as shown in FIG. 4, and the weight is changed according to the distance from the center line of the overlapping area.

A block diagram of the embodiment described above is shown in FIG.
Shown in In FIG. 5, 501 is an input terminal, 502 is an input HDTV signal, 503 is an image division unit, 504 is an overlap area setting unit, 505 and 506 are divided small images, 50
7, 508 are discrete cosine transform units, 509, 510 are quantization units, 511, 512 are code assignment units, 513, 5
14 is a buffer memory unit, 515 and 516 are encoded outputs for small images, 517 is a multiplexing unit, and 518 is HDTV.
Coded data for a signal, 519 is a coded data output terminal, 520 is a coded data input terminal, and 521 is an input H.
DTV signal encoded data, 522 is a demultiplexing unit, 52
3, 524 is encoded data of the separated small image, 52
5, 526 is a code decoding unit, 527, 528 are inverse quantization units, 529, 530 are inverse discrete cosine transform units, 531.
Reference numeral 532 is an overlap determination unit, 533, 534 is an overlap-decoded image for a small image, 535 is data indicating a pixel position in an overlap region, 536, 537 are overlap-decoded images for a small image after weighting coefficient processing, and 538 is a weight setting unit. , 539, 540
Is a non-overlapping area of each small image, 541 is a weighted average processed image of overlapping decoded images, 542 is an image combining unit, 543 is a final HDTV decoded image, and 544 is an output terminal.

First, the HD input from the input terminal 501
The TV signal 502 is divided into two small images 505 and 506 by the image dividing unit 503. At this time, the area set by the overlapping area setting unit 504 is divided so as to be included in both of the two divided images. The divided small images 505 and 506 are input to the discrete cosine transform units 507 and 508, respectively, subjected to the discrete cosine transform, and then input to the quantization units 509 and 510. In the quantizing units 509 and 510, the discrete cosine transform coefficients are quantized, the codes are assigned in the code assigning units 511 and 512, and then sent to the buffer memory units 513 and 514.

In the buffer memory units 513 and 514, feedback control is applied to the quantizing units 509 and 510 so that the outputs thereof have a predetermined information amount. The coded outputs 515 and 516 for each small image are input to the multiplexing unit 517 to generate one coded data 518.
Then, they are sent to the transmission line through the output terminal 519.

On the other hand, on the receiving side, first, the HDTV signal coded data 521 input from the coded data input terminal 520 is separated into two small image coded data 523 and 524 by the demultiplexing unit 522. To be done.
Separated small image coded data 523, 52
4 is input to the code decoding units 525 and 526 to decode the code, respectively, and is then input to dequantization units 527 and 528, where discrete cosine transform (DCT) coefficients are reproduced and inverse discrete cosine transform unit 529, Sent to 530.

In the inverse discrete cosine transform sections 529 and 530, the reproduced DCT coefficients are subjected to inverse discrete cosine transform to obtain decoded images of the respective small images. In addition to this, the overlap determination units 531 and 532 determine that the overlap decoded image 533 is determined to be the overlap portion.
The encoded data of 534 is sent to the weight setting unit 538,
The overlapping decoded image 5 is weighted by the weighting factor w ₁ determined from the data 535 indicating the pixel position in the overlapping portion.
36,537. The duplicate decoded image 536, 537
The weighted average processed image 541 is obtained as the sum of the above, and is input to the image combining unit 542 together with the non-overlapping regions 539 and 540 of the decoded small image. Image composition unit 542
Then, these are combined as one HDTV decoded image 543 and then sent to the output terminal 544.

In the above embodiment, the case where the image is divided into two in the horizontal direction has been described. However, the case where the image is divided into a plurality of pieces in the horizontal and vertical directions can be considered in the same manner, and three or more small images are overlapped. The same idea can be extended for cases such as the following.

[0026]

As described above, according to the present invention,
In the decoded image obtained by combining the decoded small images, the weighted average of both of the small images is applied to the overlapping portion, so that it is possible to prevent the boundary line between the images from being detected.

As a result of the above, a decoded image with no discomfort can be reproduced, which greatly contributes to improving the image quality of the decoded image.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of a screen division type image encoding / decoding method according to the present invention.

FIG. 2 is a diagram showing an example of a block configuration for implementing the present invention.

FIG. 3 is a diagram illustrating an example of the present invention.

FIG. 4 is a diagram showing an example of weighting factors according to the embodiment of the present invention.

FIG. 5 is a block diagram of an embodiment of the present invention.

FIG. 6 is a diagram showing a conventional screen division type image encoding / decoding method.

FIG. 7 is a block diagram showing an example of a conventional method of parallel encoding by space division.

FIG. 8 is a diagram showing a noise level at a boundary portion of a decoded image synthesized by a conventional method.

[Explanation of symbols]

201 Input Terminal 202 Input HDTV Signal 202 Image Dividing Unit 204 Overlapping Area Setting Unit 205, 206 Divided Small Image 207, 208 Information Compressing Unit 209, 210 Encoding Output 211 Multiplexing Unit 212 Multiplexed Encoded Data 213 Output Terminal 214 Reception terminal 215 Received coded data 216 Demultiplexer 217, 218 Coded data 219, 220 for separated small image 221, 222 Coded data for decoded small image 223, 224 Duplication judgment Part 225, 226 Redundant decoded image for small image 227, 228 Non-overlapping region of each small image 229 Weighted averaging unit 230 Weighted coefficient processed image of redundant decoded image 231 Image synthesizing unit 232 Final decoded HDTV signal 233 output terminal

Claims (2)

[Claims] 1. A transmission side divides an image of a digital image signal to be encoded into a plurality of small images, and then performs high-efficiency encoding in parallel on each of the divided small images and transmits the small images. In the image coding / decoding method in which the receiving side decodes the coded data for a plurality of transmitted small images and then synthesizes the original image to obtain a plurality of small images at the transmitting side. The border area is divided so that it overlaps each other. On the receiving side, when the divided encoded and decoded small images are combined and the decoded image is reproduced, , The decoded value of the small image including the area is set as the reproduction value, and for the area where two or more small images overlap, the weighted average value for the decoded value of each overlapping small image is set. Image coding / decoding method characterized by using a reproduction value . 2. The image coding / decoding method according to claim 1, wherein a weighting factor is determined for each pixel when a weighted average is applied to the decoded values of the overlapping small images. A characteristic image encoding / decoding method.