JP2010062881A - Image encoding method and image decoding device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image compression method in which a still image is divided and moving image encoding is performed to encode the still image, and in which the efficiency of compression can be improved. <P>SOLUTION: An image encoding method for encoding a still image of digital data includes an image decomposing section for dividing a still image of a sheet using a rectangular block of N pixels×M pixels as a basic unit and generating a decomposed image obtained by collecting pixels at the same position in each divided rectangular block, and an encoding section for generating a stream by performing moving image encoding on the decomposed image. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image encoding method and an image decoding apparatus.

  In recent years, video cameras and the like that can record both moving images and still images have become widespread. In such a video camera, the resolution of a still image is larger than that of a moving image, and a moving image encoder generally does not support a high-resolution still image size.

On the other hand, in a video camera device that can record both still images and moving images, having both an encoder for moving images and still images is required to be avoided because the configuration of the device becomes complicated and expensive. . Also, if the image format is different when encoding a moving image and a still image, there is a problem that file management in the apparatus becomes complicated. Therefore, the encoder is desired to be one type. When moving image encoding such as H.264 or MPEG is employed, it is necessary to divide and compress a high-resolution image due to image size restrictions.
As background art in this technical field, for example, there is JP-A-2004-120499. The gazette states that “the original image supplied from the image input unit 15 having a larger number of pixels than the HDTV image (QHDTV) is converted into a necessary number of HDTV images in the screen dividing unit 12 and the HDTV necessary for neighboring pixels of the original image. After obtaining HDTV images 13a, 13b, 13c, and 13d for masking, sampling, and compression with the same number as the number of images, encoding and decoding are performed by MPEG encoder 14 and MPEG decoder 24. In other words, it is possible to effectively obtain a good image quality even with a high-resolution moving image having a larger number of pixels than that of an HDTV image.

JP 2004-120499 A

  In the above-mentioned Patent Document 1, a method for dividing a high-resolution moving image and performing moving image coding is described. However, a method for dividing a high-resolution still image and performing moving image coding, and compression efficiency are described. It does not describe how to improve.

  The present invention provides an image encoding method that uses moving image encoding when encoding a high-resolution still image, and an image compression method that improves compression efficiency.

  The above object can be achieved by the invention described in the claims.

  According to the present invention, it is possible to provide an image compression method that uses moving image coding by dividing a still image when encoding a high-resolution still image, and an image compression method that improves compression efficiency. .

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  As shown in FIG. 1, the video camera device 100 representing the first embodiment of the present invention includes an imaging unit 101, an image decomposition unit 102, a buffer unit 103, an encoding control unit 104, and an encoding unit. 105, a recording medium control unit 106, and a recording medium unit 107.

  The imaging unit 101 has a function of capturing both still images and moving images as digital data. In particular, it is assumed that the imaging unit 101 can capture a high-resolution still image. The image decomposition unit 102 generates a plurality of decomposed images 151 from the still image 150 captured by the imaging unit, as will be described later. The buffer unit 103 is a storage area for temporarily buffering the decomposed image.

  The encoding control unit 104 receives the decomposition information 152 from the image decomposition unit 102, and performs control 153 of the encoding unit 104 and control 154 of the recording medium control unit 105 based on the decomposition information 152. The encoding unit 105 reads the decomposed image 155 from the buffer, performs moving image encoding, and generates a stream 156. The recording medium control unit 106 records the stream 156 and the disassembly information 152 on the recording medium 107 as recording data 157.

  In the above description, the video camera apparatus is taken as an example of the image processing apparatus. However, the image encoding method of the present embodiment is another apparatus that can handle both still images and moving images, for example, digital television reception. The present invention can also be applied to devices such as a still camera equipped with a video encoder. Further, the image encoding may be performed by the CPU executing the image encoding program without using dedicated hardware by a personal computer or the like. Furthermore, an apparatus that can handle only still images may be used instead of an apparatus that can handle both still images and moving images.

  The image decomposition method of the image decomposition unit will be described with reference to a still image 200 of 2560 × 1920 pixels in FIG. The still image 200 is divided into rectangular blocks of 2 pixels × 2 pixels, and pixels at the same position (same number) in the rectangular block are collected to generate a decomposed image (1280 × 960 pixels) of 201 to 204. .

For the sake of explanation, each image shown in FIG. 2 represents only a part of a pixel. Further, although the still image 200 of 2560 × 1920 pixels is used, any image size may be used. Furthermore, the rectangular block of 2 pixels × 2 pixels may be an arbitrary rectangular block of N pixels × M pixels. In that case, N × M decomposed images are generated.
In the example of FIG. 2, the disassembly information 152 is N = 2 and M = 2. A rule such as numbering from the upper left to the raster scan order as in (1) to (4) of the high resolution still image 200 in FIG. 200 can be decrypted. This disassembly information 152 may be embedded as additional data in the stream or may be recorded as separate information. Further, the decomposition information may be eliminated by predetermining N and M by the encoding device and the decoding device.

  FIG. 3 shows an example of decomposition by the decomposition method. For example, when the still image 300 is decomposed by the decomposition method, decomposed images 301 to 304 having a high correlation between images are generated. When the decomposed images 301 to 304 are encoded as moving images, since the correlation between frames is high, inter-frame prediction can be performed efficiently. As a result, the compression efficiency is improved.

FIG. 4 shows picture types and coding order when moving image coding is performed. MPEG and H.264 For example, H.264 video coding is based on an I picture (Intra coded picture) that is coded only by prediction within a screen, and a P picture (Predictive coded picture) that is coded in a forward direction between frames. ) And a B picture (Bidirectionally coded predictive picture) that performs bidirectional prediction in the past and the future and encodes the difference. Although FIG. 4 shows an example of the encoding result of the decomposed images 301 to 304, one or more I pictures and any number of P pictures and B pictures may be used.
Here, FIG. 5 shows an example of the code amount when encoding is performed using JPEG (still image encoding) and MPEG2 (moving image encoding). In general, in MPEG2, the ratio of the code amount for making all the pictures of I, P, and B pictures have the same image quality is such that the code amount of P picture is 10, Is said to have a code amount of 5. Assuming that the MPEG2 I picture and JPEG have substantially the same compression rate, as shown in FIG. 4, the total code amount when the I picture 301, the P picture 302, the B picture 303, and the B picture 304 are sequentially encoded. The code amount 501 is twice the code amount of the I picture by adding all of the code amount ratios. On the other hand, the total code amount 500 when the still image 300 is encoded by JPEG is equal to the case where all the decomposed images 301 to 304 are encoded by I pictures, and is therefore four times the code amount of I pictures. Therefore, when encoded with MPEG2, the code amount can be halved compared to when encoded with JPEG.
According to the example described above, an apparatus that handles both a moving image and a still image only needs to have a moving image encoding device, and the cost can be reduced. In addition, the compression efficiency can be improved when encoding a still image.

  In general, when encoding a P picture or a B picture, a wide search range such as ± 64 pixels is searched in order to obtain an optimal motion vector. In the present invention, since similar pictures exist at almost the same positions between frames, the search range is limited to a narrow range such as ± 1 pixel, so that encoding efficiency equivalent to a wide range of search can be maintained. However, the motion vector search process can be greatly reduced, and the power consumption during encoding can be reduced.

A case where lossless encoding is used for moving image encoding will be described with reference to FIG. In lossy encoding, since a decoded image is used for inter-frame prediction, a plurality of decomposed images cannot be encoded simultaneously. However, when performing lossless encoding in which data is completely restored, since the decoded image and the original image match, it is possible to perform inter-frame prediction using the original image as a reference image instead of the decoded image. It is possible to simultaneously encode the separated images. By simultaneously encoding, the processing time can be reduced.
In the above description, the case where all the decomposed images 301 to 304 are encoded at the same time has been described. However, any number of images may be encoded at the same time, such as encoding two decomposed images simultaneously. In addition, in the case of lossless encoding of not only a decomposed image but also a moving image, a plurality of images can be encoded simultaneously by using the original image as a reference image.

  FIG. 7 shows an image decoding apparatus 600, which includes a recording medium unit 601, a decoding unit 602, an image composition control unit 603, an image composition unit 604, and an image adjustment unit 605.

In the recording medium unit 601, a stream and disassembly information 650 are recorded. The decoding unit 602 decodes the stream 651 recorded on the recording medium and creates a decoded image 655. The image composition control unit 603 performs control 652 of the decoding unit 602, control 653 of the image composition unit 604, and control 654 of the image adjustment unit 605 based on the disassembly information 650 recorded on the recording medium. The image composition unit 604 creates one still image 656 from a plurality of decoded images based on the decomposition information.
The image adjustment unit 605 creates a display image 657 with the image size adjusted by matching the still image 656 created by the image composition unit 604 with the size of the display unit. The display unit 606 displays the display image 657.

A still image synthesizing method of the image synthesizing unit will be described with reference to FIG. First, the pixels 201 having the same coordinates are collected from the decoded images 201 to 204 based on the decomposition information to generate a rectangular block of N pixels × M pixels. Next, the original high-definition still image 200 can be restored by arranging the rectangular blocks in the order of the coordinates of the collected pixels. Here, to arrange in the coordinate order, the rectangular block that collects the pixels of coordinates (1, 1) is the upper left end of the image, and the rectangular block that collects the pixels of coordinates (2, 1) is (1, 1) It means that a rectangular block in which pixels of coordinates (1, 2) are collected to the right of the rectangular block is arranged below the rectangular block (1, 1).
An example of a decoding device linked with the display unit will be described with reference to FIG. In general, when the display size is small, for example, when displaying thumbnails, all high-resolution images are decoded and then reduced to the display size. In the present invention, since the decomposed image is a reduced image of the still image before encoding, all images are not decoded such as decoding only I pictures, decoding I pictures, and P pictures according to the display size. Even a reduced image can be obtained. Therefore, decoding is performed until the image size closest to the display size 700 is obtained. For example, when the decoded image 701 of the I picture has an image size closest to the display size, the image is enlarged to the display size. If the decoded picture of the I picture is larger than the display size, it is reduced to the display size.

  With the above processing, the processing load during decoding can be greatly reduced, and power consumption can be reduced and decoding speed can be improved. Further, the display speed of thumbnails is improved by improving the decoding speed.

  Although only the decoding device is described in FIG. 7, the decoding device in FIG. 7 may be included in the image processing device in FIG.

It is an example of the video camera apparatus of a 1st Example. It is an example of an image decomposition method. It is an example of a divided image. It is an example of the encoding method. It is an example of the amount of codes when encoded by JPEG and MPEG2. This is an example of speeding up processing when lossless encoding is used. It is an example of an image decoding apparatus. It is an example of the decoding apparatus cooperated with the display part.

Explanation of symbols

100 video camera equipment
101 Imaging unit
102 Image division
103 Buffer
104 Coding control unit
105 Encoder
106 Recording medium controller

Claims (9)

In an image encoding method for encoding a still image of digital data,
An image decomposing unit that divides one still image by using a rectangular block of N pixels × M pixels as a basic unit, and generates a decomposed image in which pixels at the same position in each divided rectangular block are collected;
An encoding unit that encodes the decomposed image to generate a stream;
An image encoding method comprising: 2. The image encoding method according to claim 1, wherein the moving image encoding uses inter-frame prediction. The image encoding method according to claim 1, wherein the encoding unit has a function of inputting and encoding a moving image.   3. The image encoding method according to claim 2, wherein the encoding unit has a function of inputting and encoding a moving image. 5. The image encoding method according to claim 4, wherein when the decomposed image is encoded as a moving image, a range for searching for a motion vector is limited to a narrow range compared to a case where a moving image is input and encoded. An image encoding method characterized by the above. When using lossless encoding in an image encoding method that encodes a plurality of images of digital data, it is possible to perform encoding processing simultaneously by using the original image as a reference image for inter-frame prediction. A characteristic image encoding method.   2. The image encoding method according to claim 1, wherein when lossless encoding is used for moving image encoding, the original image is used as a reference image for inter-frame prediction, and encoding processing is performed simultaneously. An image encoding method to be performed.   A decoding unit that decodes a stream encoded by the image encoding method according to claim 1, and collects pixels at the same coordinates of the decoded decomposed image to generate a rectangular block of N pixels × M pixels, and collects An image synthesizing apparatus comprising: an image composition unit that restores one still image by arranging the rectangular blocks in the order of pixel coordinates. A decoding unit that decodes a stream encoded by the image encoding method according to claim 1, and collects pixels at the same coordinates of the decoded decomposed image to generate and collect a rectangular block of N pixels × M pixels An image composition unit that restores one still image by arranging the rectangular blocks in pixel coordinate order;
An image decoding apparatus comprising: an image adjusting unit that enlarges / reduces an image size according to a display size; and changing the number of decoded images decoded by a decoding unit according to the display size.

Images