JPH01307382A - Encoding control system - Google Patents

Abstract

PURPOSE:To suppress the generation of an effective block and to decrease an information generating quantity by not evenly decreasing a block discrimination threshold when a picture stops but making it variable in units of a block in a block discrimination part. CONSTITUTION:The block discrimination part consists of a threshold control part 201, a discrimination control part 1, a memory 3, and a block decision part 5. Here, when the picture tops, the block discrimination threshold is not evenly decreased to a desired value in units of a frame, but the threshold of the whole frames is decreased to the desired value after fixed time passage, and, from the viewpoint of a space, by gradually expanding the area, where the threshold is decreased to the desired value, in stages. Thus, when the block discrimination threshold is decreased, the sharp increase of the number of the effective blocks can be suppressed, and the information generating quantity can be decreased.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a coding control method in an interframe coding device that highly efficiently codes an input image signal sequence.

[Prior art] Figure 6 shows, for example, 1985 Technical Report C885 of the Institute of Electronics and Communication Engineers.
-3 Murakami, Ito, Kamizawa "Color moving image transmission system for video conferencing" is a block diagram showing the configuration of a 7V-to-frame encoding device using a conventional encoding control system, p.
102 is a subtracter that performs subtraction between the input image signal sequence 101 and the previous frame reproduced image signal sequence 112 and outputs an inter-frame difference signal sequence 103; 106 is an encoding/decoding unit that subtracts the input image signal sequence 101 and the previous frame reproduced image signal sequence 112; It inputs identification data 105 and outputs encoded data 107 and decoded interframe difference signal sequence 108.

Further, 109 is the previous frame reproduced image signal series 11.
2 and the decoded frame difference signal 108 and outputs the reproduced image signal sequence 110. Reference numeral 111 is a frame memory that stores t-1 frames of the reproduced image signal sequence 110.

113 is a transmission buffer that inputs the encoded data 107, calculates the amount of information generated, and outputs an encoded control signal 114 and transmission data 115.0116 inputs the encoded control signal 114 and performs encoding control. Parameter 117
This is an encoding control unit that outputs the block identification unit 104 to the block identification unit 104.

Reference numeral 104 denotes a block identification unit which inputs the encoding control parameter 117 and the interframe difference signal sequence 103 and outputs a block identification data 105.

FIG. 7 shows a configuration example of the block identification unit 104 in FIG. 6.

201 is a threshold control unit that outputs a threshold value Th202t associated with the encoding control parameter 117;
03 is a block determination unit that performs diblock determination using a threshold value Th 202.

Next, the operation will be explained.

The input image signal sequence 101 is converted into an inter-frame difference signal sequence 103 by subtracting the previous frame reproduced image signal sequence 112 by a subtracter 102 . Since this inter-frame difference signal sequence has a lower overall signal power than the original signal, encoding with a smaller encoding error is possible with the same encoding lid.

This inter-frame difference signal sequence 103 is encoded and decoded by a
06, high-efficiency encoding/decoding is performed, and encoded data 1
07 and a decoded interframe difference signal sequence 108 are obtained.

An adder 109 adds the previous inter-frame difference signal sequence 112 and the decoded inter-frame difference signal sequence 108 to obtain a reproduced image signal sequence 110.

This reproduced image signal sequence 110 is stored in the frame memory m and delayed by a predetermined frame time to form a previous frame reproduced image signal sequence for encoding the next frame, while the encoded data 107 is stored in the transmission buffer 113. After being stored, it is sent out as transmission data 115.

In addition, the transmission buffer 113 obtains the sum of the code amounts for one frame as an encoding control signal 114 (information generation amount) and supplies it to the encoding control section 116.The encoding control section 116
The encoding control buff meter 117 used in the block identification unit 104 is based on this information generation ji 114 and an encoding mode signal such as a fixedly selected encoding speed and reproduction image quality based on an external instruction. The block identification unit 104 adaptively controls the encoding control parameter 11 for each frame.
In order to judge the block of the entire frame with the threshold [Th] associated with 7, the threshold value control unit 201 outputs Th 202, and the comparison between Th and the data value is performed to judge the block according to the following formula. The block identification data 105 is determined by the block identification data 105.

Data value 4 Th ν=O invalid block> Th
ν=1 Valid block For this valid block, the data value corresponding to the block is output as encoded data 107 together with block identification data ν105, and for invalid blocks, the data value corresponding to the block is assumed to be O. The coding control method used in conventional interframe coding devices is executed as described above, so when the input image is stationary, the block identification threshold is uniform for each frame. If the number of blocks is lowered, the number of effective blocks per frame increases, resulting in an extremely large amount of generated information.

This invention was made to solve the problems of the first generation, and it is possible to suppress the sudden increase in the number of effective blocks when lowering the block identification threshold, and reduce the amount of information generated. The purpose is to obtain a coding control method.

[Means for Solving the Problems 1] The encoding control method according to the present invention does not uniformly lower the block identification threshold to a desired value on a frame-by-frame basis when an image becomes still, but , after a certain period of time,
In addition, spatially, the threshold value of the entire frame is lowered to a desired value while expanding the region in which the threshold value is gradually lowered to a desired value in stages.

[Function J] The encoding control method according to the present invention makes the block identification threshold variable within a frame, so that it is possible to suppress a rapid increase in the number of effective blocks and an extreme increase in the amount of information generated.

[Embodiment J Hereinafter, one embodiment of the present invention will be described with reference to the drawings. 1st
The figure shows an example of the configuration of a block identification unit for executing the encoding control method invented by Jin. Reference numeral 201 indicates a threshold f! associated with the encoding control buff meter 117. Th
202 is the threshold value Th output from the threshold value control section; 1 is the threshold value To for comparison, and the diagonal line in FIG. Input the value of the threshold f [T11T 2 s T 3 s T 4 and the buff meter S4 indicating the area for each area of area 1, area 2, area 3, and area 4 shown in , and set the threshold value Th
In the current frame, the threshold value TI −T2 for each region to be used is
&'ra This is an identification control unit that determines and outputs the value of 'ra, a comparison threshold TO, and a buff meter S2 indicating the area. 3 is a comparison threshold i]iI'.

and the threshold value T1%T2%T3 for each area. This memory stores the buff meter 8 indicating the value and area of T4 and updates it every frame. 5 is a threshold value T1 for each area used in the current frame output from the identification control unit l.
This is a block determination unit that performs block determination based on the value of s Th % Tm s T4 and outputs block identification data ν105'i.

This FIG. 1 corresponds to the block identification section 104 in FIG. 7, so other parts are the same as FIG. 7.

Next, the operation will be explained.

When the image is stationary, if the block identification threshold is uniformly lowered for each frame in the block identification section, the number of effective blocks for which the value of block identification data ν is 1 increases,
The amount of information generated increases rapidly as shown by the solid line in FIG.

Therefore, in order to prevent the amount of information generation from increasing rapidly, the comparison threshold value To represents the threshold value of the previous frame when Th is larger than TO, and otherwise represents the desired value. The threshold value is compared with the threshold value Th 202 output from the threshold value control unit 201, and if Th is smaller than TO, the threshold value of the entire frame is set to the desired value at once as shown by the solid line in FIG. Rather than lowering it to a value, gradually,
In order to make the threshold for the entire frame reach the desired value after a certain period of time has passed, the frame in which Th changes from not smaller than To to smaller is the first frame, the next frame is the second frame, and the next frame is the second frame. If the frame is the third frame and the next frame is the fourth frame, then the first frame is the third frame and the next frame is the fourth frame. Second. Third. Step 4: Lower the threshold value gradually as shown by the dotted line in Figure 3, so that it approaches the desired value.
When the processing of the fourth frame is completed, processing is performed so that the threshold value of the entire frame becomes a desired value.

The following explanation will be based on FIG. The identification control unit 1 shows the magnitude relationship and area between the threshold Th202 associated with the encoding control parameter 117 output from the threshold control unit 201 and the comparison threshold To read out from the memory 3. Depending on the parameter SO value, the fast value T1%T2%TB% for each area used by the block determination unit 5
At the same time as updating the value of T4, the value of the buff meter S indicating the comparison threshold TO% area stored in the memory 3 is updated, and the threshold values Ts, T2 %T3, T for each area are updated.
4, the comparison threshold value To, and the value of the buff meter S indicating the area. In the block determination unit 5, the threshold value T1 s T2 for each area output from the identification control unit l,
Tm % Block determination is made at Tm, and block identification data ν105 is output.

The following description will be made with reference to FIG. 5, which shows the flow of processing based on the configuration example of the block identification section in FIG. 1.

First, the threshold Th202 associated with the encoding control buff meter 117 output from the threshold control unit 201
The comparison threshold value TO read out from the memory 3 is compared in magnitude. Next, the magnitude relationship between Th and To is that Th is 1°
If the value is not greater than 0, the value of the puff meter 8 indicating the area read from the memory 3 is checked. Depending on the magnitude relationship between Th and TO and the value of S, the following processing is performed.

(0) When the θth stage Th is larger than TO, and when Th is equal to T and the directness of S is 0, the identification control unit 1 calculates TO% Tl s 'i"2, T3 % T
All values of m are updated to Th, and the value of S is output as 0 and stored in the memory 3.

The block determination unit 5 determines the entire frame as a block using the threshold value Th, and outputs block identification data ν105.

(1) When the first stage Th is smaller than To and the value of B is 0, the identification control unit 1 updates the value of s'ro%T1 to Th and the value of day to IK, and updates T2, The value of T3 s Tm is output as is and stored in the memory 3.

In the block determination unit 5, the area 1 is determined as T+=Th=T.

So, region 2.3.4 is the same as the previous stage T2- Ts , T
Block determination is made using m, and block identification data ν105 is output.

(2) When the value of the second stage S is 1, the identification control unit 1 updates the value of T2 to T, and the value of S to 2, 'ro s T 1%T3, Tm The value "" is output until that time and stored in the memory 3.The block determination unit 5 determines the area 1+2 as a block with To and the area 3.4 with the same T S and T 4 as in the previous step, and stores the block identification data. Output ν105 (3) When the value of S in the third stage is 2, the color is identified. In the control unit 1, the value of T3 is updated to TO, the value of S is updated to 3, and To %Tl %T2 %T4 The block determination unit 5 outputs the value as it is and stores it in the memory 3.The block determination unit 5 determines the area 1 + 2 + 3 as a block with 're, and the area 4 with 4 as in the previous stage, and uses the block identification data ν10.
(4) When the value of the fourth stage S is 3, the identification control unit 1 updates the value of Tm to To and the value of S to OK, and the values of To s Tt , T2 s Ts are It is output as is and stored in the memory 3.

In the block determination unit 5, the entire frame is determined as a block using the threshold value To, and the processing in the first, second, third, and fourth stages of outputting the block identification data ν105 is performed when Ttl - T is observed first. Assuming that the threshold value Tht−T in the frame is
In (3), the area 1 + 2 + 3, and in (4), the width value of the entire frame is lowered to T.

In this example, we set the area to spread radially,
Also, the time it takes for the threshold value of the entire frame to reach the desired value is 4 frames, but this is just an example. The time required for the drop to drop is also completely arbitrary.

For example, it is also possible to set the area randomly at m periods (m is a natural number) to cover the entire frame.

In addition, the interframe encoding device shown in FIG. 7 performs intraframe encoding on areas that are sequentially specified according to a predetermined order synchronized in transmission and reception, and encodes the areas so that the entire frame is intraframe encoded after a certain frame time has elapsed. The present invention can also be applied to a mode in which a frame is set based on the above embodiment, sequence information indicating the frame time elapsed is transmitted, and areas other than the specified area are interframe encoded. .

[Effects of the Invention j] The nine-frame intercoding device using the coding control method according to the present invention has a block identification section that does not uniformly lower the block identification threshold when the image is still, but By changing the angle in units, it is possible to suppress the generation of effective blocks, which has the effect of reducing the amount of information generated.

Furthermore, if the image begins to move while the threshold value is being lowered, the amount of information generated will not increase rapidly, so it goes without saying that tracking of the movement will be better.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing an example of the configuration of the block identification section of the uninvented encoding control method, Fig. 2 is a diagram showing the relationship between the threshold value and the amount of information generated, and Fig. 3 is the threshold value. FIG. 4 is a diagram showing an example of area setting, FIG. 5 is a diagram showing the flow of processing in the block identification section, and FIG. 6 is a diagram showing the flow of processing in the block identification section of the conventional block identification section. FIG. 7 is a block diagram showing the structure of the interframe encoding device shown in FIG. 6. In the figure, 117 is an encoding control buff meter, 201 is a threshold control section, 202 is a threshold Th, 1 is an identification control section, 21 is a threshold value T 1 , T2 , T3 for each area used in the current frame, Parameter 8.3 indicating T4 and comparison threshold To and area is memory, 4 is threshold 'I' 1% for each area used in the previous frame
T2, T, %T4, comparison threshold value To, and buff meters 8 and 5 indicating the area are block determination units, and 105 is block identification data. In addition, in the figures, the same reference numerals indicate the same or equivalent parts.

Claims (2)

[Claims] (1) Control the block identification threshold for executing the conditional pixel replenishment on a frame-by-frame basis in an interframe encoding device that highly efficiently encodes an input image signal sequence using conditional pixel replenishment on a block-by-block basis. In the process of setting the new block identification threshold to a desired value smaller than the block identification threshold used in the past, the new block identification threshold is , a threshold value that is set to a value that decreases stepwise, and is finally set to the desired value and applied to the entire frame. (2) The intra-frame area to which the new block identification threshold is applied is expanded to areas that are sequentially specified according to a predetermined order synchronized in transmission and reception, and this area is intra-frame encoded, The area is set in frame units so that the entire frame is intra-frame encoded after a certain frame time has elapsed, and sequence information indicating the elapse of the frame time is transmitted for each frame, and areas other than the specified area are Means for periodically performing an intra-frame/inter-frame mixed coding mode for performing inter-frame coding, and adaptively controlling the intra-frame/inter-frame mixed coding mode and the inter-frame coding mode using threshold control. A coding control method characterized by switching to.