JPH0262179A - System for transmitting motion quantity of picture transmission equipment - Google Patents

Abstract

PURPOSE:To decrease an information quantity to be transmitted by obtaining deviation from the initial deflection of the motion quantity of a notable block in a set retrieve range. CONSTITUTION:An initial deflection arithmetic means 1 obtains the initial deflection of the motion quantity based on prescribed algorithm from the detected motion quantity of the circumferential block of the notable block. Based on the initial deflection, a retrieve range setting means 2 sets the detection range smaller than the motion detection maximum range. In the range, the deviation of the motion quantity from the initial deflection is obtained by a deviation arithmetic means 3, and sends the value to a decoder. In the decoder, based on the detected motion quantity of the circumferential block of the notable block, the initial deflection of the motion quantity of the notable block is obtained by an initial deflection arithmetic means 4. The initial deflection and the received deviation are synthesized by a motion quantity arithmetic means 5, and the motion quantity of the notable block is obtained. Thus, the information quantity to be transmitted can be decreased.

Description

DETAILED DESCRIPTION OF THE INVENTION [Summary] The present invention relates to a motion amount transmission method in an image transmission device such as a high-efficiency encoding device for moving images.

The purpose is to further reduce the amount of motion information to be transmitted.

On the sending side, the initial deviation of the amount of motion of the block of interest is determined based on a predetermined algorithm from the amount of detected motion of blocks surrounding the block of interest, and based on this initial deviation, a search range smaller than the maximum motion detection range is determined. is set, the deviation of the amount of motion from the initial deviation within this search range is determined, and this deviation information is sent to the receiving side.The receiving side calculates the movement of the block of interest based on the detected motion amount of the surrounding blocks of the block of interest. It is configured to obtain an initial deviation of the quantity, and to combine this initial deviation with the received deviation to obtain a motion counterfeit of the block of interest.

[Industrial application field]

The present invention relates to a motion amount transmission method in an image transmission apparatus such as a high-efficiency encoding apparatus for moving images.

In recent image transmission, including TV conference telephone systems, various methods have been proposed from the viewpoint of compressing the amount of transmitted information as much as possible and reducing the bit rate in order to make effective use of transmission paths and reduce terminal costs. has been done.

Among these, the so-called block matching motion compensation prediction method has been attracting attention as this type of information compression method. In this method.

By detecting the movement of the transmission screen and compensating the predicted screen by the amount of movement to obtain a prediction error (difference signal), the amount of prediction error is further reduced and transmitted, thereby achieving an even lower bit rate. ing.

In such a motion compensated predictive coding method, the purpose is to further reduce the bit rate of transmitted information.

In addition to reducing the signal amount of prediction errors, it is necessary to further reduce the signal amount of motion information to be transmitted.

[Conventional technology]

Conventionally, as a motion amount encoding method in a motion compensation predictive encoding method, there is a first-order method.

Namely.

(1) When the amount of movement in an image is small, the frequency of occurrence is high, so Huffman encoding is used to encode the image with a variable length depending on the amount of occurrence. (2) For images where the amount of movement is very often zero. In the case of , the combination method % expression % uses this to perform run-length encoding when the amount of motion is zero, and performs Huffman encoding according to the frequency of occurrence in other cases.Here, the amount of motion information to be transmitted is: Block of interest to be encoded (hereinafter referred to as block of interest)
A maximum range within which the target block is expected to move is set, and motion vector information indicating the movement position of the block of interest within this maximum range is used.

[Problem to be solved by the invention]

In the conventional motion amount encoding method, (1), (2)
In either case, it is necessary to transmit an amount of information corresponding to the maximum detection range in which motion is to be detected, and it is necessary to use the code length assigned to the code corresponding to this maximum detection range. As a result, the code length becomes longer, which increases the circuit scale for encoding.Also, when a screen with a large amount of motion continues, the amount of code to be transmitted increases rapidly, and the amount of motion information is not correctly transmitted. It may also happen that the transmission becomes impossible.

Therefore, it is an object of the present invention to further reduce the amount of motion information to be transmitted.

(Means for solving problems) FIG. 1 is a principle block diagram according to the present invention.

In FIG. 1, there are nine image encoding apparatuses according to the present invention.

Initial deviation calculation means for calculating the initial deviation of the amount of motion of the block of interest based on a predetermined algorithm from the detected amount of motion of surrounding blocks of the block of interest.1. A search range setting means 2 for setting a search range smaller than the maximum motion detection range based on this initial deviation; and a deviation calculation means 3 for calculating the deviation of the amount of movement from the initial deviation within this search range. , and is configured to transmit the deviation information obtained by the deviation calculating means 3 to the receiving side.

Also, pictures related to the present invention@! The if encoding device includes an initial deviation calculation means 4 for calculating an initial deviation of the amount of motion of the block of interest based on the detected amount of motion of blocks surrounding the block of interest. and,
It is equipped with a motion amount calculation means 5 which combines this initial deviation and the received deviation to obtain the amount of motion of the block of interest.

[Effect]

In the image encoding device, the initial deviation calculation means 1 calculates the initial deviation of the amount of motion of the block of interest from the detected amount of motion of blocks surrounding the block of interest based on a predetermined algorithm. The search range setting means 2 sets a search range smaller than the maximum motion detection range based on this initial deviation. Further, the deviation calculation means 3 calculates the deviation of the amount of motion from the initial deviation within this search range, and sends this deviation information to the decoding device.

On the decoding device side, the initial deviation calculating means 4 calculates the initial deviation of the amount of motion of the block of interest based on the detected motion amounts of the blocks surrounding the block of interest.

This initial deviation and the received deviation are combined by the motion amount calculation means 5 to obtain the motion amount of the block of interest.

〔Example〕

The present invention will be described in detail below with reference to one drawing.

An image encoding device as an embodiment of the present invention is shown in FIG. In this embodiment, a motion compensation predictive coding circuit of a motion detection type having an initial deviation and a motion amount coding section are shown.

In FIG. 2, the subtracter 11 inputs one block of input screen
The difference between this and the motion compensation predicted screen intersection output from the variable delay unit 17 is taken and output as a prediction error screen ε. A quantizer 12 quantizes the prediction error screen ε. A Huffman encoding circuit 13 performs Huffman encoding on the output q of the quantizer 12.

An adder 14 adds the output q of the quantizer 12 and the predicted screen intersection to generate locally decoded double-sided X'. Frame memory 1
5 stores a previous screen obtained by delaying the locally decoded screen X' of the adder 14 by about one frame.

Does the variable delay device 17 move the previous screen of the frame memory 15?
ili compensator 16 to generate a motion compensated predicted screen intersection.

The motion compensator 16 receives the input screen X and the previous screen of the frame memory 15, and detects the amount of motion (motion vector Vo) of the input screen X based on these inputs.

The detected motion vector Vo is sent to the variable delay unit 17, and each inserted motion amount information is sent to the motion amount encoding section 8. This motion amount information includes 1-bit motion amount information B1 for determining whether or not the motion is zero. There is motion deviation information Δ■ and 1-bit specific motion amount selection information B2 for determining whether or not a specific motion amount has been selected.

Movement? The motion amount information output from the i-compensator 16 is encoded by a motion amount encoder 18 and output. The motion amount encoding unit 18 includes a run-length encoding circuit 181 and a Huffman encoding circuit 182.

The run-length encoding circuit 181 performs run-length encoding on the frequently occurring motion amount information B1 and sends it to the decoder side. On the other hand, the Huffman encoding circuit 182 performs Huffman encoding on the motion deviation information Δ■ which can take a plurality of different values and the specific motion amount selection information B2 as one group, and sends the Huffman encoding to the decoder side.

The motion detection operation in the motion compensator 16 will be described in detail below. First, for a block of interest X, which is a human-powered screen that is currently being focused on to be encoded, an initial deviation vector VI is determined in accordance with a predetermined algorithm, based on the detected motion vectors of surrounding blocks that are already adjacent to the code ratio. For example, as shown in FIG. 4, the motion vector v~ of the block M at the same position in the previous screen with respect to the block of interest X, the motion vector VW of the block directly above the block of interest X in the current screen, or the movement Among the zero vectors V (0, O), the motion vector with the smallest cumulative error sum of the block of interest X after the deviation is set as the initial deviation vector (1).

After the initial deviation vector (1) is determined, a secondary deviation vector ΔV is searched within a predetermined search range E.

Determine the optimal motion vector Vo. Namely.

As shown in FIG. 5, a predetermined search range E is set around block X1, which is the target block X displaced according to the initial deviation vector V+, and while block X+ is moved within this search range E, The block Xn with the smallest cumulative error sum is found by checking, and the direction vector from the block XI to the block XD is set as the deviation vector Δ■. In this case, the optimal motion vector Vo of the block of interest X is the vector sum of the initial deviation vector V+ and the deviation vector Δ■. Namely.

Vo=V++Δ■ It is.

Here, the search range E is set to be sufficiently smaller than the maximum motion detection range E+*ax of the block of interest X. In this embodiment, the search range E is ±3 pixels vertically and horizontally, the maximum detection range Eax is ±15 pixels vertically and horizontally, and the number of blocks for matching within the search range is 49. As a result, the number of reference blocks in the search range E becomes sufficiently smaller than the number of reference blocks in the maximum detection range Ewbax.

Here, if the motion vector V of the block directly above the block of interest is selected as the motion vector Vo, the bit of the specific motion amount selection information B2 is set to "l", and if the motion amount is zero, The bit of motion information B1 is set to "1".

A block configuration of an image decoding device as an embodiment of the present invention is shown in FIG. This embodiment shows a motion compensation predictive decoding circuit of a motion detection type having an initial deviation and a motion amount decoding section.

In FIG. 3, a Huffman IM encoding circuit 21 performs Huffman decoding on the quantized output (A). The adder 22 adds the output signal of the Huffman decoding circuit 2I and the predicted signal intersection from the variable delay device 24 to restore the decoded signal X'. The frame memory 23 gives the decoded signal X' a delay of one frame and sends it to the variable delay unit 24. The variable delay device 24 generates a motion compensated predicted screen intersection by shifting the previous screen based on the motion vector Vo determined on the receiving side independently of the transmitting side.

The run length decoding circuit 25 receives the transmitted signal (B
), the motion amount information B1 is run-length decoded and applied to the control input terminal Pc of the switch circuit 20. The Huffman decoding circuit 2G performs Huffman decoding on the motion deviation information ΔV and the specific motion amount selection information B2 from the transmitted signal (C), and supplies the deviation information Δ■ to the adder 27 and the specific motion amount selection information B2. are respectively input to the control input terminal Pc of the switch circuit 29.

The motion memory 28 delays the final motion vector Vn given to the variable delay unit 24 by one frame time, for example, and outputs it to the addition section 327 as an initial deviation vector v1, and also delays it by one block line to obtain a specific motion vector. V
It is output to the input terminal pl of the switch circuit 29 as ll.

The adder 27 receives the deviation vector ΔV from the Huffman decoding circuit 26 and the initial deviation vector V+ from the motion memory 28.
are added and inputted to the input terminal P2 of the switch circuit 29 as the final motion vector Vo.

The output signal from the switch circuit 29 is input to the input terminal P1 of the switch circuit 20. The input terminal P2 of the switch circuit 20 has Vo corresponding to a motion vector with zero motion.
A signal of (0,0) is input.

The output signal from this switch circuit 20 is transmitted to a variable delay device 24.
and is also input to the motion memory 28.

These switch circuits 28 and 29 are connected to the control input terminal Pc.
When it receives the O'' signal, it goes to the input terminal Pl side.

On the other hand, when it receives the "l" signal, it switches to the input terminal P2 side.

The operation of this decoding circuit will be explained. When both the motion amount information B and the specific motion selection information B2 are “O”, the deviation vector ΔV decoded by the Huffman decoding circuit 26 is added to the initial deviation vector V+ read from the motion memory 28 in addition 327. are added vectorially to generate a motion vector Vo. This motion vector Vo is transferred to the variable delay device 24 via the switch circuit 29 and the switch circuit 20.
is inputted into the frame memory 23 to shift the previous picture in the frame memory 23 and generate a motion-compensated predicted picture intersection.

On the other hand, if the motion amount information B1 is "1", the switch circuit 2
0 selects the zero vector Vo (0,0) and inputs it to the variable delay device 24. Further, when the specific motion amount selection information B2 is "l", the switch circuit 29
A specific motion vector Vg from 8 is selected and inputted to the variable delay device 24 via the switch circuit 20. Here, the specific motion vector VK corresponds to the motion vector VO of the block directly above the block of interest X.

〔Effect of the invention〕

As explained above, in the present invention, as movement amount information,
Search range E smaller than the amount of information in the maximum detection range Emax
This allows the amount of motion information to be transmitted to be reduced and the motion amount information to be encoded and transmitted using a small-scale circuit. This will greatly contribute to high-efficiency encoding.

[Brief explanation of the drawing]

FIG. 1 is a principle block diagram according to the present invention. FIG. 2 is a block diagram showing an image encoding device as an embodiment of the present invention. FIG. 3 is a block diagram showing an image decoding device as an embodiment of the present invention. FIG. 4 is an explanatory diagram of selection of the initial deviation vector; FIG. 5 is an explanatory diagram of the search range for the optimal motion vector Vo. In fig. 11-Subtractor 12-Quantizer 13.182--Huffman encoding circuit 14.22.27--Adder 15.23-Frame memory 16-Motion compensator 17.24-Variable delay unit 18-- Motion amount encoding unit 181 - Run length encoding circuit 1.26 - Huffman decoding circuit 5 - Run length decoding circuit 8 - Motion memory 0.29 - m - Switch circuit Principles related to the Section! Sugiyoshi 1 free time) Most ■ Pe 7
To 1 Shino Nana City ρ 8 Socket Moon Figure 5

Claims (1)

[Claims] 1. On the transmitting side, the initial deviation of the amount of motion of the block of interest is determined based on a predetermined algorithm from the amount of detected motion of blocks surrounding the block of interest, and the maximum motion detection range is determined based on this initial deviation. Set a search range smaller than , find the deviation of the amount of motion from the initial deviation within this search range, send this deviation information to the receiving side, and the receiving side calculates the detected movement of blocks surrounding the block of interest. A motion amount transmission method for an image transmission device configured to obtain an initial deviation of the amount of motion of a block of interest based on the amount of motion, and to combine this initial deviation and a received deviation to obtain the amount of motion of the block of interest. 2. Initial deviation calculation means (1) for calculating an initial deviation of the amount of motion of the block of interest based on a predetermined algorithm from the amount of detected movement of surrounding blocks of the block of interest; A search range setting means (2) for setting a small search range, and a deviation calculation means (3) for calculating a deviation of the amount of movement from an initial deviation within this search range, the deviation calculation means ( An image encoding device configured to transmit the deviation information obtained in step 3) to a receiving side. 3. Initial deviation calculation means (4) that calculates an initial deviation of the amount of motion of the block of interest based on the detected amount of movement of surrounding blocks of the block of interest; An image decoding device comprising a motion amount calculation means (5) for calculating the motion amount of a block.