JPH0448874A - Image data recording method - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Field of Industrial Application] This invention provides a method for compressing and encoding still image data and recording it on a recording medium so that the amount of code after compression encoding is less than or equal to a predetermined target amount of code. The present invention relates to an image data recording method that guarantees the number of images that can be recorded on a recording medium with a fixed storage capacity by compressing the images into a predetermined number of images and then recording them, and is particularly suitable for application to electronic still cameras.

[Prior art] A method that combines orthogonal transformation with variable length coding is considered to be effective as a highly efficient compression coding technology for natural images (still images), and this method is also included in the international standard for color natural image coding. is scheduled to be adopted (Journal of the Institute of Image Electronics Engineers: V
OR18, No. 6, P398-P407).

In this type of encoding system, the quality of the decoded image and the amount of output code after encoding can be controlled by controlling certain parameters. The relationship between the amount of code and image quality is that the larger the amount of code, that is, the lower the degree of compression, the smaller the deterioration of the image quality from the original image. The deterioration of image quality will be significant. However, this type of encoding method performs adaptive processing using the local correlation of images during encoding, making it possible to perform highly efficient compression encoding according to the image quality of the target image. There is.

[Problem to be solved by the invention]

Incidentally, in general, an image with a detailed pattern has a large amount of code, and a flat image with many solid areas has a small amount of code.

Therefore, even if compression encoding is performed using the same code amount control parameter, the obtained code amount will differ for each target image.

Therefore, with this type of encoding method, it is difficult to predict in advance the amount of code generated by compression encoding. It is not always easy to do, especially
Until now, applications of natural image coding have mainly been considered for image transmission such as still image videophones and facsimiles, and storage in image databases, so encoding must be performed to obtain a certain amount of code. was not necessarily required.

However, when applying this type of encoding method to, for example, an electronic still camera, it is necessary to guarantee the number of images that can be recorded on the recording medium. A technique for compressing and encoding a target image with a constant code amount is required.

This invention changes the code amount control parameter for each target image in order to guarantee the number of images that can be recorded when compressing and encoding multiple still image data and recording it on a recording medium with a fixed storage capacity. An object of the present invention is to provide an image data recording method in which image data of each target image is compressed to a predetermined target code amount or less and then recorded on a recording medium.

[Means to solve the problem]

The present invention provides an image data recording method in which input image data for one screen is compressed and encoded using a method that combines orthogonal transformation and variable length encoding, and then recorded on a recording medium. Compression encoding is performed multiple times, the resulting code amount is measured and compared with the predetermined target code amount for one screen, and based on the comparison result, the code amount after compression encoding of the input image data is determined. A code amount control parameter that is equal to or less than the target code amount is determined, and the input image data is compressed and encoded using the determined code amount control parameter and recorded on the recording medium.

[For production]

According to this invention, when compressing and encoding a target image and recording it on a recording medium, trial encoding is performed on the target image in advance, and the resulting code amount and a predetermined target code amount per screen are determined. From the comparison result, find a code amount control parameter that can make the code amount after compression encoding of the target image less than or equal to the target code amount, and use this determined parameter to perform the actual compression of the target image. Encoding is performed and the compressed data is recorded on a recording medium. In this way, it is possible to reliably guarantee the number of still images that can be recorded on the recording medium.

〔Example〕

FIG. 1 is a block diagram showing an embodiment of an image data recording method according to the present invention, and shows an example in which the present invention is applied to an electronic still camera.

In FIG. 1, input image data captured from an imaging unit (not shown) is stored in a frame memory l for each screen, divided into a plurality of blocks each having 8×8 pixels, and then sent to a discrete cosine transform circuit 2. is supplied to The conversion circuit 2 performs two-dimensional DCT (Discrete CT) for each block.
Co51ne Transforms) and supplied to the quantum northbound W13. The 8×8 DCT coefficients F, J (i, j-0, l, ..., 7) obtained by the conversion circuit 2 represent components obtained by decomposing one block of input image data into spatial frequencies. , coefficient F of coefficient F =j. represents a value (DC component) proportional to the average value of input image data, and as the variable 1yJ increases, the higher frequency component (
AC component).

The DCT coefficients F 2 obtained in this way are linearly quantized in a quantization circuit 3 with different quantization step widths for each coefficient.

The quantization step width is determined by adding a coefficient 2" (n=o, ±1.±2.
) is specified by the value multiplied by in this case,
The width n of the coefficient 2'' is called a scaling factor and is supplied from the scaling factor control circuit 6. An example of the quantization matrix is shown in FIG.

Each coefficient of the DCT coefficient F ij quantized by the quantization circuit 3 is scanned from a low-order coefficient to a high-order coefficient by a zigzag scan circuit 7, and converted into one-dimensional data.
" run length value and valid value (value other than "0") are converted into two-dimensional data. Next, Huffman encoding is performed in a Huffman encoding circuit 8, and the recording medium 9 such as an IC card is stored as variable length encoded data. An example of a zigzag scan table is shown in FIG.

Incidentally, the amount of code finally obtained from the Huffman code circuit 8 as variable-length coded data can be controlled by changing the scaling factor n.

That is, by increasing the scaling factor n, the quantization step width can be increased and the amount of code obtained can be reduced; conversely, by increasing the scaling factor n
By reducing , it is possible to reduce the quantization step width and increase the amount of code obtained. Although this qualitative relationship generally holds regardless of the target image, the quantitative relationship between the scaling factor n and the amount of data after compression is not uniquely determined and differs depending on the target image. . Therefore, even if the target image is encoded with the same scaling factor n, the amount of code obtained will differ depending on the content of the image.

Therefore, in this embodiment, a code amount measuring circuit 10 is provided on the output side of the Huffman encoding circuit 8, and the image data of the target image stored in the frame memory l is A trial encoding is performed on the data, the resulting code amount is measured by the code amount measuring circuit 10, and the measurement result is supplied to the scaling factor control circuit 6. The control circuit 6 compares the code amount supplied from the code amount measurement circuit IO with a predetermined target code amount allocated to one screen, and determines from the comparison result that the code amount after compression encoding is less than or equal to the target code amount. Find the scaling factor n. Then, actual compression encoding of the target image is performed using the obtained scaling factor n, and the compressed data is recorded on the recording medium 9.

In this case, the attempted encoding for the target image is 100 m5 from the start of acquisition of the target image to the actual start of acquisition.
Assuming that a delay of about ec is acceptable, and if the compression encoding process of the target image can be done in about 3Q+wsec, then as shown in Fig. 4 (
It can be done three times as shown in a) and (b).

That is, in FIG. 4(a), capture of the target image is started at time To, and scaling factor no (
The first trial encoding is performed using the initial value).

As a result, if the code amount after compression encoding is larger than the target code amount, the scaling factor control circuit 6 increases the scaling factor no, and at time T1 a new scaling factor n1 (=no+Δ)(Δ>0 )
A second trial encoding is performed. In this way, trial encoding is repeated by repeatedly adding Δ to the scaling factor flQ to increase the quantization step width until the code amount after compression becomes equal to or less than the target code amount. Then, when the code amount after compression becomes less than the target code amount for the first time, the target image is actually compressed and encoded using the scaling factor at that time, and the result is recorded on the recording medium 9. In the example of FIG. 4(a), the code amount becomes less than the target code amount for the first time at time T2, so actual encoding is performed at time Ta using the scaling factor n2 (-n o +2 x Δ) at this time and the recording medium is Start accumulating to 9.

Furthermore, as shown in Fig. 4, etc., the scaling factor no.
When the first trial encoding was performed using (initial value), if the output code amount was smaller than the target code amount, the scaling factor no was decreased and a new scaling factor n1 (=n(l-Δ) The second trial encoding is performed at time T1 by .In this way, the scaling factor n is increased until the output code amount exceeds the target code amount.
Iteratively subtracts Δ from , decreases the quantization step width, and repeats trial encoding. Then, actual compression encoding of the target image is performed using the scaling factor immediately before the output code amount exceeds the target code amount for the first time, and the result is recorded on the recording medium 9. In the example of FIG. 4(b), actual encoding is performed at time T3 using the scaling factor nl (=no - Δ) at time T1 immediately before time T2 when the target code amount is exceeded for the first time, and the data is stored on the recording medium 9. Start accumulating.

In this way, before starting the encoding accumulation of the target image, trial encoding is performed on the target image multiple times, and the resulting output code amount is measured and compared with the target code amount.
The scale factor n is increased or decreased, and actual encoding and accumulation is performed using the scaling factor n when the output code amount exceeds or falls below the target code amount for the first time. In this way, it is possible to reliably compress and encode the target image with a code amount that is less than or equal to a predetermined target code amount.

In the above embodiment, DCT was used as the encoding method, but other encoding methods such as vector quantization, predictive encoding such as DPCM, or block encoding may also be used. It is applicable to all image encoding systems that use variable length codes.

Further, as the code amount control parameter for controlling the output code amount, the quantization matrix itself may be directly changed in addition to the above-mentioned scale factor. In this case, since each coefficient in the DCT coefficient matrix can be changed individually, more accurate code amount control can be performed.

[Effects of the Invention] According to the present invention, when a plurality of still images are encoded in real time using a method that combines orthogonal transformation and variable length encoding and recorded on a recording medium, the target still images are The code amount is calculated in advance by encoding, the code amount control parameter that is less than the target code amount is determined from the result, and the target image is compressed and encoded using the control parameter. It becomes possible to reliably guarantee the number of still images that can be recorded by a simple means.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of the image data compression recording method according to the present invention, FIG. 2 is a diagram showing a numerical example of a quantization matrix, FIG. 3 is a diagram showing the order of zigzag scan, and FIG. is a diagram for explaining the time relationship of the present invention. Quantization matrix of luminance signal Fig. 2 Zigzag scan table Fig. 3

Claims (2)

[Claims] (1) In an image data recording method in which input image data for one screen is compressed and encoded using a method that combines orthogonal transformation and variable length encoding, and then recorded on a recording medium, multiple The amount of code obtained is measured and compared with the predetermined target amount of code per screen, and based on the comparison result, the amount of code after compression encoding of the input image data is An image data recording method characterized in that a code amount control parameter that is equal to or less than a target code amount is determined, and the input image data is compressed and encoded using the determined code amount control parameter and recorded on the recording medium. (2) The image data recording method according to scope item 1, wherein the code amount control parameter is a scaling factor.