JP2001103483A - Moving picture coding method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a moving picture compression system, that can enhance moving picture compression efficiency without losing the advantages of the Motion JPEG system, such as commonality with still picture compression, ease of decoding of an optical frame, each of edit and a small circuit scale. SOLUTION: This moving picture coding method applies JPEG compression to each frame and is characterized in that the JPEG compression is applied to a specific frame, after it is converted into difference information with respect to a preceding reference frame thereto.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a moving picture encoding method for JPEG-compressing inter-frame difference information.

[0002]

2. Description of the Related Art As a moving image compression method in a digital camera or the like, a so-called Motion JPEG compression method in which each frame of a moving image is JPEG-compressed as a still image is widely used.

[0003] The Motion JPEG compression method, which regards a moving image as a plurality of still images and compresses each with the JPEG compression method, is convenient as a moving image that can be easily compressed.

On the other hand, in the MPEG system for compressing a video signal composed of a moving image, the video signal is encoded using a line correlation or an inter-frame correlation of the video signal.

[0005] For example, in the intra-frame encoding process, the time t = t1, t2, t3,...
When the images PC1, PC2, PC3,... Constituting a moving image are to be transmitted, the image data to be transmitted is subjected to orthogonal transform processing such as DCT (Discrete Cosine Transform) processing, compressed, and transmitted. Things.

[0006] The inter-frame coding process uses an inter-frame correlation of a video signal to sequentially process adjacent picture PCs 1.
, And the image data PC12, PC23,..., Which are pixel data differences between PC2, PC2 and PC3,.

As a result, the video signal system is changed to an image PC
1, PC2, PC3,... Have a smaller data amount as compared with the case of reproducing all the image data.

[0008]

SUMMARY OF THE INVENTION However, Motion JPE
The G method has the commonality with still image compression, has the advantage of being able to decode random frames at random, is easy to edit, and has the advantage of a small circuit scale, but has the problem of low compression efficiency. .

On the other hand, moving picture coding methods such as MPEG have high compression efficiency, do not have commonality with still picture compression, and have difficulty in decoding arbitrary frames at random.
Furthermore, there is a problem that the editing is not easy and the circuit scale is increased.

[0010]

The invention according to claim 1 is
In a moving image encoding method for JPEG-compressing each frame of a moving image, a specific frame is converted into difference information from a preceding frame and then JPEG-compressed.

According to a second aspect of the present invention, the frame to be converted into the difference information is within a predetermined frame interval from a frame which is not converted into the difference information (hereinafter referred to as “reference frame”), and the average value of the difference , The total value or the maximum value is equal to or less than a predetermined amount.

[0012] In the invention according to claim 3, the difference information is:
After subtracting the pixel value at the same position of the reference frame from each pixel value of the frame to be converted into the difference information, adding a predetermined amount, the process falls within the range of pixel values assumed by JPEG compression, and then JPEG compression is performed. It is characterized by doing.

[0013] In the invention according to claim 4, the difference information is:
JPEG compression is performed by setting a high-frequency quantization width smaller than that of the reference frame.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below in detail with reference to the drawings.

First, an outline of the encoding process and the decoding process will be described with reference to FIGS.

FIG. 1 is a block diagram of the encoding process according to the present invention.

As shown in FIG. 1, input frame image data 101
Is the JPEG compressor 10 according to the judgment of the switching judgment unit 105.
9, the input frame image data 101 is directly JPEG-compressed, or after the input frame image data 101 is converted by the difference converter 103 into a difference from the reference frame image data held in the frame memory 104,
JPEG compression is performed by adding the switching information 108, and encoded data 111 is obtained. Here, the “reference frame image” refers to a frame image directly encoded immediately before.

FIG. 2 is a block diagram of a decoding process according to the present invention. As shown in FIG. 2, the encoded data 111
It is decoded by the JPEG decoder 201. Next, in accordance with the added switching information 110, the restored frame data 208 may be used as it is, or the decoded frame image data 202 may be added to the reference frame image data 206 held in the frame memory 205 by the difference restorer 204. To obtain the restored frame image data 208.

Next, details of the processing will be described with reference to FIGS.

FIG. 3 is a flowchart of the encoding process. First,
The input frame number and the reference frame number are initialized to 1 (301), and image data for one frame is input (3
02). As described above, the input image data is directly JPEG-encoded, or converted into difference information and JPEG-encoded, and the first determination criterion (303) is whether the input frame number is the first frame or the reference frame. When the frame interval between the frames becomes a predetermined value, JPEG encoding is directly performed (31).
0).

As shown in FIGS. 1 and 2, the difference frame is encoded and decoded based on the data of the reference frame held in the frame memory. This is image data, and becomes decoded image data on the decoding side. For this reason, in the image data encoded as the difference frame, an error due to the difference in the contents of the reference frame occurs in addition to the encoding error. Therefore, by directly encoding at regular intervals, accumulation of errors is prevented.

If the above condition (303) is not satisfied, the difference from the reference frame is evaluated (304). The value of the input frame is F [i], the value of the reference frame is G [i], i =
Assuming that 1, .., N (N is the total number of pixels), the average e of the absolute differences between the two is obtained.

[0023]

(Equation 1)

If this value is smaller than a predetermined threshold value, it is encoded as a difference frame, and if it is greater than or equal to the threshold value, the input data is directly encoded (305). this is,
This is because, when the value of e is large, if encoding is performed as a difference frame, an encoding error may appear as a pseudo contour on the restored image.

If the average e of the difference absolute values is smaller than the threshold value, the difference from the reference frame is calculated. The input frame value is F [i], the reference frame value is G [i], i = 1, .., N
When (N is the total number of pixels), generally, the value of the difference frame is D [i],

[0026]

(Equation 2)

Where the value range of the above equation is from -255 to 25
5 and not compatible with the JPEG input range. However, the range of the input image data is from 0 to 255. Therefore, in the present invention, an offset of 128 is added by the following equation,

[0028]

(Equation 3)

Further, the value is rounded to the range of 0 to 255 (3
06).

[0030]

(Equation 4)

Thus, the range of the difference data is adjusted to the range of 0 to 255 which is the input range of JPEG.

The obtained difference image data D [i] is encoded by JPEG (308). At this time, the value of the quantization width is set to a value smaller than the value of the quantization width in the encoding of the reference frame (this is a default value), and the encoding is performed with an increased quality (307). This is because, for a difference frame, an encoding error may appear as an error such as a pseudo contour on the restored image, and the occurrence of the error is suppressed by increasing the encoding quality.

In this method, the reference frame is a frame directly coded immediately before. This is because if the immediately preceding frame is used as the reference frame regardless of whether the frame is a directly encoded frame or a differentially encoded frame, such as MPEG, it becomes difficult to decode any frame at random. This is because the advantage of JPEG is lost.
When encoded as a differential frame, a flag indicating that the frame is a differential frame is added to the encoded data. Specifically, it is added to the user comment area in JPEG format. If necessary, a frame number of a reference frame from which a difference is obtained is also added. This is to make it possible to easily specify a reference frame when it is desired to randomly decode an arbitrary frame.

On the other hand, for the frames determined to be directly encoded as they are in the input image data in the determinations 303 and 305, the quantization width is set to a default value (309) and JPEG-encoded (310). Thereafter, the input frame is updated as a new reference frame (311).

The input frame number is incremented, and the above processing is performed again on the next frame. The coding is completed by repeating this for all frames.

Next, the details of the decoding process will be described with reference to the flowchart of FIG. Input frame number and reference frame number are 1
(401), one frame of encoded data is input (402), and it is JPEG decoded (40).
3). When a flag indicating a difference frame is added to the user comment area of the decoded data,
This is passed to the difference restoration processing (405). If not,
The decoded image data is directly used as the restored image data (40
6), the restored image data is set as new reference frame image data (407).

The data passed to the difference restoration processing is such that the decoded frame image data is D [i] and the reference frame image data is G
When [i], i = 1,..., N (N is the total number of pixels), the frame data is restored to the restored frame image data F [i] as follows.

[0038]

(Equation 5)

The input frame number is incremented, and the above processing is performed again on the next frame. By repeating this for all frames, decoding is completed.

Here, an example for demonstrating the present invention will be described below. When an original image is obtained by converting a 30-frame image from the beginning of the ITU-T standard image “Foreman” into a monochrome image of 176 × 144 size, according to the encoding of the present invention, the data amount is 132,911 bytes, average SN
The ratio was 36.92 dB.

On the other hand, each frame is JPEG encoded (Mo
data, the data amount is 145,82
0 bytes, and the average SN ratio was 36.82 dB.

Thus, although the average SN ratio is almost equal, the data amount in the encoding of the present invention is reduced, and the encoding efficiency is improved.

[0043]

As is apparent from the above description, according to the present invention, commonality with still image compression, easy decoding of an arbitrary frame, easy editing, and a reduction in circuit scale are achieved. The moving image compression efficiency can be improved without impairing the advantages of the Motion JPEG method.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an encoding process according to the present invention.

FIG. 2 is a block diagram illustrating a decoding process according to the present invention.

FIG. 3 is a flowchart showing an encoding process of the present invention.

FIG. 4 is a flowchart showing a decoding process of the present invention.

Claims (4)

[Claims] 1. A moving image encoding method for JPEG-compressing each frame of a moving image, wherein a specific frame is converted into difference information from a preceding frame and then JPEG-compressed. . 2. A frame to be converted into the difference information,
Frames not converted to difference information (hereinafter “reference frame”)
That. ), The frame satisfies both or one of the following conditions: within a certain frame interval, and an average value, a total value, or a maximum value of the differences is equal to or less than a predetermined amount. 2. The moving picture encoding method according to 1. 3. The difference information is obtained by subtracting a pixel value at the same position of the reference frame from each pixel value of a frame to be converted into difference information, and adding a predetermined amount to the pixel value assumed by JPEG compression. JPEG after processing within the range
2. The moving picture coding method according to claim 1, wherein the moving picture is compressed. 4. The moving picture coding method according to claim 1, wherein the difference information is JPEG-compressed by setting a high-frequency quantization width smaller than that of the reference frame.