JP2942097B2 - Special moving object image extraction device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to the extraction of a special moving object image for extracting a person or the like reflected on a security camera, or a special moving object image for monitoring a vehicle or the like reflected on a traffic monitoring camera. More specifically, the apparatus uses a motion-compensated inter-frame prediction to transmit a luminance signal component to a non-moving part, and transmits a luminance signal component and a chrominance signal component to a moving part. The present invention relates to a special moving object image extracting device capable of reducing the amount of information and extracting a moving image by color.

[0002]

2. Description of the Related Art In recent years, with the development and spread of digital transmission lines and advances in image processing technology, effective use of image information services is expected due to the development of high-speed digital signal processing and LSI technology therefor. In particular, in the division of communication, research on high-performance networks capable of efficiently transmitting image information has been actively pursued.
It is expected that a very high quality image transmission service represented by N (Integrated Services Digital Network) will be provided.
Further, it is expected that image information is efficiently used in fields other than communication.

Generally, the amount of information contained in image information is very large, and if an image signal is handled as it is, a considerably high transmission rate results, which is not practical. However, the information amount of the image signal includes much redundancy, and by removing this redundancy, the information amount can be reduced.
Therefore, studies on an image coding method for that purpose are being actively conducted. The image signal includes temporal information such as movement and change and spatial information relating to the content of one image frame, and each has redundancy. In particular, in an image (moving image) signal, the correlation between consecutive frames is strong, and the amount of information can be significantly reduced by reducing temporal redundancy. As a coding method utilizing such a temporal correlation, a motion compensation frame (field) inter-prediction coding method is widely and generally used.

FIGS. 3A and 3B are diagrams for explaining the principle of a motion-compensated predictive coding system, which is one of the conventional image coding systems. In the motion compensated prediction coding method,
The image frame to be encoded and the previously encoded image frame are each divided into appropriate rectangular block areas, and comparison is performed for each of the block areas. A motion vector indicating a direction in which a matching block is located in a previously encoded image frame from a block located at the same position as the target block is detected in a previously encoded image frame. The motion vector is used to represent motion information included in the image signal, and a matching block in the video frame is set as a predicted value for the target block area.

FIG. 4 is a block diagram of a conventional image coding apparatus. In the figure, reference numeral 31 denotes a motion compensation inter-frame prediction unit, 32 denotes a frame memory unit, 33 denotes a difference calculation unit, 34 denotes an error coding unit, and 35 Is an error decoding unit, and 36 is an addition unit. The input image signal is sent to the motion compensation prediction unit 31 in a rectangular area unit that does not overlap for each image frame. The motion compensation prediction unit 31 performs motion compensation prediction between the input image signal and an image signal input before the input image signal stored in the frame memory unit 32, and detects and outputs a motion vector. You. According to the result of the motion compensation prediction process,
The area determined to be most appropriate from the frame memory unit 32 is read as a predicted value. The difference between the predicted value and the input image signal is obtained by the difference calculation unit 33,
A prediction signal is output. The prediction coded signal is sent to the error coding unit 34. The error encoding unit 34 uses the correlation of the prediction error signal in the image frame to obtain a DCT (Disc).
Appropriate encoding such as rete Cosine Transform (discrete cosine transform) encoding is performed. The result is output to the outside of the device together with the motion vector from the motion compensation prediction unit 31 as a coded signal.

On the other hand, the encoded signal is locally decoded by an error decoding section 35, and a prediction error decoded signal is output.
The prediction error decoded signal is again added to the prediction value read from the frame memory unit 32 in the addition unit 36, and a reproduced image signal after encoding is obtained. The reproduced image signal is sent to and stored in the frame memory unit 32. When encoding of one image frame is completed,
The frame memory unit 32 stores a locally decoded image signal of the image frame, and is used as a prediction value of motion compensation prediction of the next encoding target image frame. Separate encoding, which separates a decoded signal into a luminance signal and a chrominance signal and encodes them separately, is used to handle the chrominance signal. This separation coding is also called component coding, and is adopted in coding at a low transmission rate because information can be easily compressed. FIG. 5 shows the mechanism of separation encoding. That is, the input image signal is converted into a luminance signal (Y) component, a color signal (r) component, and a color signal (C
b) separated into components.

[0007]

As described above, when image information is transmitted at a low transmission rate, a luminance signal (Y) and a chrominance signal (Cr.C) are used even if motion compensation inter-frame prediction is used.
Since the signal of b) is transmitted, the amount of basic information is large.
For this reason, when the motion is large, a phenomenon called an overflow that exceeds the limit of error coding in motion compensation inter-frame prediction occurs. Therefore, the movement of an object having a strong movement is not transmitted as a smooth movement. In addition, in a simple color image, there is a problem that it is not possible to clearly distinguish a moving object or the like from a non-moving background or the like.

[0008] The present invention has been made in view of such circumstances, and reduces the amount of information to be transmitted, makes the movement of a moving object image smoother, and allows the moving object image to be focused on. It is intended to provide an extraction device.

[0009]

In order to achieve the above object, the present invention provides an image signal separating section for separating an input image into a luminance signal (Y) and a chrominance signal (Cr.Cb); and the color signal control unit for controlling the transmission of color signals by the information, the
Switch means controlled on / off by the color signal control unit
When a coding unit for each encoding the luminance signal and the chrominance signal, and a decoding unit for decoding the encoded signal, detects motion, and motion-compensated interframe prediction unit for transmitting the motion vector, A frame memory for storing the previous frame;
An error encoding unit for encoding the difference between the previous frame image and the current frame image, and an error decoding unit for decoding the difference ,
In the case of an image having no image, the switch means is turned off and a luminance signal is obtained.
To reduce the amount of information, and
Turn on the switch means to transmit the luminance signal and the chrominance signal
It is characterized by the following.

[0010]

In the moving object image extracting apparatus according to the present invention, when an image signal is input, the image signal is separated into a luminance signal (Y) and a chrominance signal (Cr.Cb) by an image signal separating unit. When no motion is detected by the motion compensation inter-frame prediction unit, the color signal control unit outputs only the encoded luminance signal information without outputting the separated and encoded color signal information. . In the case of a moving object image in which a motion is detected by the motion compensation inter-frame prediction unit, color signal information is output by the color signal control unit. In other words, when transmitting an image, only the luminance signal component is transmitted to a portion where there is no motion, and a luminance signal component and a color signal component are transmitted to a portion where motion is detected. A moving object is extracted by paying attention to the moving object. In addition, since the amount of information as a whole is reduced, smoother tracking image information of the moving object is transmitted even at a low transmission rate.

[0011]

Embodiments will be described below with reference to the drawings. FIG. 1 is a block diagram for explaining an embodiment of a special moving object image extracting apparatus according to the present invention, in which 11 is an image signal separating section, 12 is a luminance signal error encoding section, and 13 is a chrominance signal error. An encoding unit, 14 is a luminance signal error decoding unit, 15 is a chrominance signal error decoding unit, 16 is a chrominance signal control unit, 17 is a chrominance signal frame memory unit, 18 is a luminance signal frame memory unit, and 19 is a motion compensation frame. The inter prediction unit, 20, 21 are difference operation units, 22, 23 are addition units, and 24 is a switch.

An input image signal is supplied to an image signal separation unit 1.
1 separates into a luminance signal (Y) and a chrominance signal (Cr, Cb). The separated luminance signal (Y) and color signal (Cr,
In the case of Cb), differences from the previous frame image are calculated by the difference calculation units 20 and 21, respectively, to obtain prediction error signals. The respective prediction error signals are input to the luminance signal error encoding unit 12 and the chrominance signal error encoding unit 13, respectively, to obtain encoded signals. Each of the encoded signals is input to a luminance signal error decoding unit 14 and a chrominance signal error decoding unit 15, respectively. The decoded signals for obtaining the decoded signals are added with predicted values by adding units 22 and 23, respectively. Are stored in the color signal frame memory 17 and the luminance signal frame memory 18, respectively.

On the other hand, the input image signal is input to a motion compensation inter-frame prediction unit 19, which detects a motion and transmits a motion vector. The motion compensation inter-frame prediction unit 19
Is input to the color signal control unit 16, and the switch 24 connected to the color signal error encoding unit 13 is turned on / off by the color transmission control signal from the color signal control unit 16 so that the color signal Output information.

That is, if an image signal is input, the image signal separation unit 11 separates the image signal into a luminance signal (Y) component and a chrominance signal (Cr · Cb) component. Each is subjected to DCT (discrete cosine transform), quantized, coded, and transmitted. A part is stored in the frame memory units 17 and 18. In the first image, no motion is detected by the motion compensation inter-frame prediction unit 19 because the previous image frames are not stored in the frame memory units 17 and 18, and therefore, the switch 24 is turned off by the color signal control unit 16. Therefore, only the encoded luminance signal (Y) component is output. For this reason, a monochrome image is output only from the luminance signal component, and the amount of transmitted information is reduced.

Subsequently, when an image signal including a moving object image is input, motion is detected by the motion compensation inter-frame prediction unit 19, and a motion vector is transmitted. At this time, since the motion information is transmitted, the color signal control unit 16 turns on the switch 24 of the image portion corresponding to the motion information, and transmits the error of the encoded color signal. For this reason, only the moving object image is colored, and the information of the portion where the motion is detected is transmitted, and the information of the moving object color image is transmitted similarly to the function of the conventional motion compensation inter-frame prediction 19. to continue.

FIG. 2 is a flowchart for explaining the operation of the special moving object image extracting apparatus according to the present invention. Hereinafter, the steps will be sequentially described. First, an image signal is input (step 1), and the image signal is separated into a luminance signal and a chrominance signal (step 2). Next, the motion compensation inter-frame prediction unit performs a motion compensation process (step 3), and detects a motion vector (step 4). Next, a difference calculation is performed on the separated luminance signal and color signal to obtain a prediction error signal (step 5), and an encoding process is performed to obtain an encoded signal (step 6). In step 4, color signal transmission control is performed based on the detected motion information (step 7). The st
After the processing of ep6 and step 7, the encoding result is output (step 8), and the process returns to step 1. The luminance signal and chrominance signal encoded in step 6 are decoded (step 6).
9), add it to the predicted value (step 10), store it in the frame memory unit (step 11), and return to step 3.
In this way, the amount of information as a whole is further reduced, and only the moving object image is colored, so that the moving object image can be noticed and the moving image can be extracted by the color.

[0017]

As is apparent from the above description, according to the present invention, an image without motion transmits only a luminance signal, and an image with motion transmits a luminance signal and a chrominance signal. The amount can be reduced, and thus image communication can be performed on a communication network with a low transmission rate. Further, by using the mobile object image extraction device of the present invention, it is possible to extract a moving object even at a long distance by using a low transmission rate network.

[Brief description of the drawings]

FIG. 1 is a configuration diagram for explaining an embodiment of a special moving object image extraction device according to the present invention.

FIG. 2 is a flowchart illustrating an operation of the special moving object image extracting apparatus according to the present invention.

FIG. 3 is an explanatory diagram for explaining the principle of conventional motion compensated prediction coding.

FIG. 4 is a configuration diagram of a conventional image encoding device.

FIG. 5 is a configuration diagram of an image signal separation device in conventional separation coding.

[Explanation of symbols]

11 image signal separating unit, 12 luminance signal error encoding unit,
13: color signal error encoding unit, 14: luminance signal error decoding unit, 15: color signal error decoding unit, 16: color signal control unit,
17: chrominance signal frame memory unit, 18: luminance signal frame memory unit, 19: motion compensation inter-frame prediction unit, 20,
21: difference operation unit, 22, 23: addition unit, 24: switch.

Claims (1)

(57) [Claims] And 1. A video signal separating unit for separating an input image into a luminance signal and a color signal, a color signal control unit for controlling the transmission of color signals by the motion information, OH by the color signal controller
On / off control means , an encoding unit for encoding the luminance signal and the chrominance signal, a decoding unit for decoding the encoded signal, and detecting motion and transmitting a motion vector. includes a prediction unit motion compensation inter-frame, and the frame memory unit for storing the previous frame, and error encoding unit for encoding a difference between the previous frame image and the current frame image, the error decoding unit for decoding a difference, motion The image without
Switch off to reduce the amount of information by transmitting only the luminance signal.
For moving images, turn on the switch means.
And transmitting a luminance signal and a chrominance signal .