JPH03247091A - Moving picture differential coding device - Google Patents

Abstract

PURPOSE:To make a frame rate constant independently of the movement of picture data by allowing a bit width command section to command a bit width reducing input picture element data to a bit width variable section based on the information quantity per unit time of a differential coding data detected from a buffer memory. CONSTITUTION:A bit width command section 4 commands a bit width reducing an input picture element data to a bit width variable section 1 based on the information quantity per unit time of a differential coding data detected from a buffer memory 3. The information quantity per unit time stored in the buffer memory 3 is detected and the information quantity is compared with the information quantity per preceding unit time and when a current unit time of information quantity is increased, the section 4 commands to increase the reduction of bit number to the bit width variable section 1 and when the current unit time of information quantity is decreased, the section 4 commands to decrease the reduction of bit number to the bit width variable section 1. Thus, when the information quantity subjected to differential coding between frames is increased, the bit number per picture element is decreased to average the information quantity for each frame thereby making the frame rate constant.

Description

[Detailed Description of the Invention] [Summary] A moving picture differential encoding device that differentially encodes image information of a digital moving picture and sends the differentially encoded image information to a storage device maintains a constant frame rate regardless of the amount of movement of image data. The object of the present invention is to realize a video differential encoding device that outputs to a storage device, and includes a bit width instruction section, a bit width variable section that reduces the bit width of input pixel data and outputs it under the control of the bit width instruction section; a differential encoding unit that has a bit width matching unit that corrects the number of bits of a predicted value in accordance with the number of bits of pixel data input from the bit width variable unit, and that generates and outputs differential code data between frames; a breadthreader memory that stores and outputs the differential code data output from the differential encoder; Based on this, the bit width variable unit is configured to instruct the bit width to be reduced in the input pixel data.

(Industrial Application Field) The present invention relates to a moving picture differential encoding device that differentially encodes image information of a digital moving picture and sends the differentially encoded image information to a storage device.

In recent years, with the demand for multimedia computer systems, there has been a demand for computer systems to handle moving images as digital image information. For this reason, devices are provided that convert analog image information such as video signals into digital image information, but since the amount of information in digital images is extremely large, it takes a long time to transfer data as it is. Since the storage speed is extremely slow when storing in a low-speed, large-capacity storage device such as a magnetic disk, a moving image differential encoding device is used that compresses the amount of information by differentially encoding this digital image information.

[Prior Art] Pixels at corresponding positions between two frames of moving image information generally have very uniform values.

A video differential encoding device compresses the amount of information by generating and transmitting information in which the difference between frames is encoded by utilizing such correlation between pixels between frames.

Differential encoding of image data by a conventional differential encoder and storage of data in an external storage device will be explained using FIG.

FIG. 8(a) is a block diagram showing the configuration of a conventional moving picture differential encoder. In the figure, 7 is a differential encoding unit;
The subtracter 1, the quantizer 72, the inverse quantizer 73, the adder 74, and the 1-frame image 75 are accumulated and the value of the pixel data of the previous frame is sent as a predicted value. The pixel data is composed of a frame memory and an encoder that encodes the values of 76 quantized pixel data.

To explain the operation of the moving image differential encoding device shown in FIG. 72 outputs quantized pixel data.

This quantized pixel data is input to the encoder 76, and is also input to the dequantizer 73, where it is dequantized.
The predicted value delayed by one frame by the frame memory 75 is added by the adder 74, and this value is input to the frame memory 75.

The predicted value output from the frame memory 75 is input to the subtracter 73 as described above.

The encoder 76 encodes the input quantized pixel data according to its value and outputs the encoded data.

This output is stored in a large-capacity storage device such as a magnetic disk that can store image data with a large amount of information.

FIG. 8(a) is a diagram showing the connection between the differential encoder 7 and the storage device 8 shown in FIG. 8(a). As mentioned above, the storage device 8 is composed of a low-speed, large-capacity storage medium such as a magnetic disk, and is connected to the differential encoder 7 via the data compressor 6.

This means that if there is no change between the current frame and the previous frame, that is, there is no movement, the difference will be zero. If there is little movement in the entire frame, the frame correlation will be high and zeros will continue after taking the difference, so by compressing the portion where the zero data continues, the amount of data can be further reduced.

[Problem H to be Solved by the Invention] When the data compressor 7 compresses the differentially encoded values as described above, when the motion is large, the correlation between frames becomes small and zeros do not continue.

In other words, the amount of information in the current frame increases, and the frame rate, which is the data transfer time for each frame, is no longer constant.

For this reason, there was a problem in that a storage device with slow access speed could not keep up with the storage capacity.

An object of the present invention is to realize a differential encoding device that outputs image data to a storage device at a constant frame rate regardless of the amount of movement of the image data.

[Means to solve the problem]

In order to solve the above problems, the video differential encoding device of the present invention includes a bit width instruction section 4, and a variable bit width that reduces the bit width of input pixel data and outputs it under the control of the bit width instruction section. and a bit width matching unit 20 that corrects the number of bits of the predicted value according to the number of bits of the pixel data input from the bit width variable unit 1, and generates and outputs differential code data between frames. and a buffer memory 3 that stores and outputs the differential code data output from the differential encoder 2, and the bit width specifying unit 4 is configured to detect the bit width from the buffer memory 3. The bit width variable unit 1 is instructed to reduce the bit width of the input pixel data based on the amount of information per unit time of the differential code data.

[Effect]

In other words, the video storage device of the present invention reduces the number of bits per pixel when the amount of differentially encoded information between frames increases, thereby averaging the amount of information for each frame and maintaining a constant frame rate. be converted into

The amount of information accumulated in the buffer memory per unit time is detected, and this amount of information is compared with the amount of information per unit time before. If the amount of information per unit time has increased, the amount of information stored in the bit width variable section is Instruct to increase the amount of bit reduction,
If the number of bits has decreased, an instruction is given to reduce the amount of reduction in the number of bits.

〔Example〕

Hereinafter, one embodiment of the present invention will be described using FIGS. 2 to 7.

FIG. 2 is a block diagram showing one embodiment of the present invention. In the figure, the same components as in FIG. 1 are indicated with the same reference numerals.

Reference numeral 5 denotes a counter, which detects the break of one frame by counting the number of pixel data. 1 is a bit that forms the bit width variable part in this embodiment.
It's a slicer.

Reference numeral 2 denotes a differential encoding unit, which includes a subtracter 21 and a quantizer 2, which are provided similarly to the differential encoder described in the above-mentioned conventional technology.
2, inverse quantizer 23, adder 24, frame memory 25
In addition to the encoder 26, a bit width matching unit 20 is provided which corrects the number of bits of the predicted value in accordance with the number of bits of newly input pixel data.

Reference numeral 3 denotes a buffer memory section which stores the data sent out from the data compression section 6 and outputs it to the storage device. Since the storage device uses a low-speed, large-capacity storage medium such as a magnetic disk, the access time is slow compared to the data transfer speed from the data compression unit.

For this reason, the buffer memory section 3 has two buffer memories, and while data is being written to one side, data previously written is read from the other side and output to the memory bag f, so that data cannot be transferred from one buffer. The method is to switch the buffers once the process is complete.

Further, 4 detects the amount of information per unit time of the differential code data stored in the buffer memory section 3, and changes the amount of information per unit time from the previous amount of information to the bit width variable section 1. This is a bit width instructing unit that instructs the bit width to be reduced in input pixel data, and specifically, a microprocessor unit (MPU) is used. This MPU also operates as a control section of the device, and controls each section of the device as described later.

The processing of image data in the video code difference apparatus of this embodiment will be described below.

FIG. 3 shows the format of image data in this embodiment. In this embodiment, it is assumed that ROB data is handled as shown in the figure. This RGB data has an independent R plane, G plane, and B plane, and each image data has an information amount of 8 bits per pixel. That is, R
The entire GB data has an information amount of 24 bits per pixel. The input RGB data is first sent in one frame only for R (8 bits), then only for G (8 bits), then only for B (8 bits), and so on for each independent plane.

Image data in the above format is first input to the counter 5, which detects the nib lane and one frame break and notifies the MPU 4 of this.

The image data output from the counter 5 is input to the bit slicer 1, where the bit width is reduced for each pixel. The width to be reduced depends on the control information (bit width designation) of the MPU 4. In this example, 24 bits (RG
The bit width of B (8 bits each) is changed to 18 bits (6 each of RGB).
bit), 15 bits (5 bits each for RGB), and 1
2 bits (R(1; B each 4 bits).

This bit width reduction is performed by cutting off the lower bits of each of the RGB data, as shown in FIG. This reduction amount is the same for one frame and does not change.

The image data whose bit width has been reduced by the bit slicer is input to the differential encoding unit 2. The operation of the differential encoding unit is basically the same as that of the conventional differential encoder, but the image data of the previous frame and the image data of the current frame stored in the frame memory 25 have different bit widths per IDk pixel. Therefore, it is not possible to simply calculate the difference.

Therefore, the frame memory 25 is
The bit width of the image data of the previous frame outputted from the image data of the current frame is matched with the bit width per pixel of the image data of the current frame, and the result is output as a predicted value. That is, Fig. 5(a)
If the number of bits per pixel of the current frame is increased as shown in Figure 5 (G), the lower bits of the predicted value are interpolated, and if the number of bits is decreased as shown in Figure 5 (g)), the lower bits are interpolated. delete.

The predicted value output from the bit width matching section in this manner is input to the subtracter 21, and the difference is calculated from the image data of the current frame. The output of the subtracter 21 that takes the difference is sent to the quantizer 22
The signal is quantized by the encoder 26, and further encoded by the encoder 26.

An example of this quantization and encoding processing is shown in FIG. As shown in the figure, the quantizer 22 has three types of quantization characteristics. This quantization characteristic is switched by the Hitokabisoto number. This quantizer is realized by a ROM, and switching is performed by addressing according to the input bit width under the control of the MPU 4, and outputs a quantized representative value for each input value. Further, this representative value is encoded by the encoder 26. This encoder 26 also encodes the input representative value using three types of encoding characteristics under the control of the MPU 4. Through this quantization and encoding processing, the amount of information per pixel is reduced in accordance with the bit width reduction amount of the bit width variable section 1.

The differential encoded data output from the differential encoder 2 is further compressed by the data compressor 6 to reduce the amount of information.

If there is no movement between the current frame and the previous frame, the difference will be zero. Therefore, the data compression unit 6 determines whether the input data is zero and how many consecutive zeros there are, and compresses the portion where zeros continue.

The data compression process by the data compression unit 6 will be explained using FIG. 7. FIG. 7(a) shows an example of the input and output of data handled by the data compression unit 6, where the input differential encoded data is 1 (per plane).
The width of each pixel is 4 bits.

Further, FIG. 7(b) is a flowchart showing data compression processing.

That is, the data compression unit 6 determines whether the input data is zero for each pixel, and if zero is detected, counts the number of zero pixels.

If it is not zero, if the previous pixel is not zero, this pixel data is output as is, and if the previous pixel is zero, the number of zero pixels counted so far is output.

The data output from the data compression section 6 is temporarily stored in the buffer memory section 3. As described above, the buffer memory section 3 has a structure having two sides of sofa memories. This is because the storage device for the output light is a low-speed, large-capacity storage medium such as a magnetic disk, so the access time is slow compared to the data transfer speed from the data compression unit 6, so data is written to one buffer memory. This is because a system is used in which previously written data is read from the other buffer memory and sent to the storage device while the data is being transferred, and when the transfer from one buffer is completed, the buffer is switched.

As mentioned above, the amount of data per frame varies depending on the bit width and the amount of motion, so the Hashifah memory section 3
When one frame's worth of data is written, the amount of data is notified to the MPU 4. The MPU 4 compares the amount of information of the current frame with the amount of information of the previous frame held, and instructs the bit width variable section 1 to make the bit width per pixel of the next frame larger or smaller than the current frame. or leave it as is. Furthermore, the MPU 4 adds data indicating how many bits per pixel the frame to be stored has been compressed to the beginning of the frame and writes it into the buffer memory 3. This is because the number of bits per pixel is required when using data stored in a storage device.

Instead of comparing with the previous frame, the bit width may be specified based on the darkness value corresponding to the information amount value of the current frame.

In this way, the bit width per pixel of the next frame is changed according to the amount of information in each frame, that is, the amount of movement.
The amount of information per frame can be made constant.

[Effects of the Invention] As explained above, according to the present invention, by making the bit width variable in advance, it is possible to keep the amount of information per frame constant.・The rate remains constant. Therefore, it becomes possible to store image data using a low-speed, large-capacity storage device, which has a great practical effect.

[Brief explanation of drawings]

Fig. 1 is a diagram explaining the principle of the present invention, Fig. 2 is a block diagram showing the configuration of an embodiment of the invention, Fig. 3 is a diagram showing the format of image data handled in this embodiment, and Fig. 4 is a bit diagram. FIG. 5 is an explanatory diagram showing a width reduction method, FIG. 5 is an explanatory diagram showing bit width matching processing, FIG. 6 is a diagram showing an example of image data quantization and encoding processing in this embodiment, and FIG. 7 is a data FIG. 8 is a diagram showing a conventional moving image differential encoding device. In the figure, 1...Bit width variable unit, 2...Differential encoding unit 3...Buffer memory unit, 4...Bit width instruction unit (MPU), 20-...
This is a bit width matching section. Principle diagram of the unexploded date and month 1st drawing I show the format of the data An explanatory diagram showing an example of the bit width matching process Σ when the bit width is matched.
(α) An explanatory diagram of the compression history of the input and output rows of the data rows. Shoulder animal

Claims (1)

[Scope of Claims] A video differential encoding device that differentially encodes an image signal of a digital video between frames, comprising: a bit width instruction section (4); and an input pixel under control of the bit width instruction section A bit width variable unit (1) that reduces the bit width of data and outputs it; and a bit width matching unit that corrects the number of bits of a predicted value according to the number of bits of pixel data input from the bit width variable unit (1). a differential encoding unit (2) that generates and outputs differential encoded data between frames; and a differential encoder (2) that stores and outputs the differential encoded data output from the differential encoder (2). a buffer memory (3), the bit width instruction section (4) includes a buffer memory (3);
) A moving picture differential encoding device characterized in that the bit width variable unit (1) is instructed to reduce the bit width of input pixel data based on the information amount per unit time of the differential encoded data detected by the bit width variable unit (1).