KR100213283B1 - Moving picture encoder - Google Patents

Abstract

The present invention relates to an apparatus for selectively encoding luminance and color signals of a moving image in the case of a large moving image. To solve this problem, an input image is implemented by a frame forming unit in a frame unit, and the implemented frame is implemented. In the apparatus for sequentially providing and encoding the difference value between the current frame and the predicted frame therefrom using DCT, quantization, VLC, etc., the frame provided from the frame forming unit and the next input of the frame Image separation unit for analyzing the encoding information on the frame based on the frame, and image control unit for controlling the encoding of the difference signal between the frame and the predicted frame based on the analysis information of the image analysis unit to determine the still image and the video In addition, there is an effect that can be efficiently encoded accordingly.

Description

Video encoder

1 is a block diagram showing a general encoder for compressing and encoding a video signal on a conventional transmitting side.

2 is a block diagram showing a video encoder according to the present invention.

* Explanation of symbols for main parts of the drawings

202: frame forming portion 204: subtractor

206: DCT 208: quantization unit

210: inverse quantization unit 212: inverse DCT

220: motion compensation unit 230: image separation unit

240: image control unit

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a video encoder, and more particularly, to a video encoder for selectively encoding luminance and color signals of a video in the case of a large video using human visual characteristics.

In general, electronic devices such as video telephones and HDTVs process video signals into digital signals. Accordingly, high rate data compression is essential to reduce the enormous amount of data involved. In particular, since a larger amount of data is generated in the video signal than in the audio signal, encoding of the video signal is an important part.

Therefore, when the transmitting side encodes the image signal using spatial and temporal correlation of the image signal and then transmits the image signal to the receiving side through the limited transmission channel, the electronic device receives the decoding.

For example, the encoder on the transmitting side removes the spatial redundancy of the video signal by using transform coding such as Discrete Cosine Transform (DCT), and further uses the differential coding through motion estimation and prediction. The video signal is efficiently compressed by eliminating temporal redundancy.

1 is a block diagram showing a general encoder for compressing and encoding a video signal at the transmitting side.

Referring to FIG. 1, an input image signal is formed in a frame unit by the frame forming unit 102 and then provided to the subtractor 104 and the motion predictor 118.

In addition, the current frame implemented by the frame forming unit 102 based on the switching operation of the switches 122-124 according to the Intra / inter operation mode includes the subtractor 104, the DCT 106, and the quantization unit 108. And then reconstructed by the inverse quantization unit 110 and the inverse DCT 112, combining the reconstructed data and the previously predicted compensated motion displacement by the adder 114, and then storing them in the frame memory 116. do.

That is, on the basis of the frame provided from the frame forming unit 102 and the frame reconstructed from the frame memory 110, the motion predictor 118 predicts a motion between two frames, and the motion displacement predicted from the motion compensator is calculated. To 120.

The motion compensator 120 provides a predicted frame made after compensating for the estimated motion displacement to the previous frame provided from the frame memory 116 to the subtractor 104 through the switch 124 and at the same time the switch 122 It is provided to the adder 114 through.

The subtractor 104 obtains the difference value based on the image frame input as the next rank and the predicted (compensated) frame provided from the motion compensator 120, and performs data DCT and quantization on the VLC (126). ) Will be provided.

On the other hand, the adder 114 combines the inverted data and the motion displacement provided through the switch 122 to the previous frame in the frame memory 116 and uses the frame for the next encoding.

The VLC 126 performs variable length coding on the data quantized by the quantization unit 108 according to the occurrence frequency of each code, with a variable length encoding having a short frequency and a long length having a low frequency. While transmitting to the decoder through 128, the buffer 128 controls the quantization unit 108 according to the input image data amount.

At this time, the decoder stores the same frame as the encoder in the frame memory, and restores the current frame to the display device by the movement displacement transmitted from the encoder.

For example, the still image can know what the screen composition is by only the object outline information on the screen composition. However, in the case of a video, the motion of the object is large so that the image can not be accurately watched (detected) for an image that cannot be detected by human vision. Could not.

In addition, by using the same DCT coefficient value and the quantization coefficient value for the still image and the video without distinction, there is a disadvantage in that it cannot be efficiently encoded.

Accordingly, the present invention has been made to solve the above shortcomings, and for an image that is too large to be detected by human vision, the characteristics of luminance (luminance) and color (chroma) signals for implementing the image are determined. Accordingly, an object of the present invention is to provide a video encoder for differentially encoding the video.

In order to achieve the above object, the present invention converts a difference signal between a current frame and a prediction frame into a DCT transform coefficient of an N × N block through discrete cosine transform, and quantizes the transformed DCT transform coefficients to a predetermined quantization step size. In the image encoder for variable length coding the quantized N × N DCT transform coefficients, a difference between the current frame and the next frame is calculated, and the current frame is moved through a comparison between the calculated difference and a predetermined threshold. Image separation means for detecting whether the image is large or small in motion and generating coded adaptive control information corresponding to the detection result; Means for generating a DCT coefficient adjustment signal for DCT and a quantization coefficient adjustment signal for quantization based on the generated encoding adaptive control information; DCT means, when the DCT coefficient adjustment signal is provided, to DCT the luminance component of the difference signal between the current frame and the predicted frame relatively tightly compared to the color component, and to DCT the rough component relatively roughly to limit the high frequency component; And when the quantization coefficient adjustment signal is provided, quantizes the DCT transform coefficients of the luminance component in a step relatively smaller than the DCT transform coefficients of the color component, and quantizes the DCT transform coefficients of the color component in a relatively large step. Means for providing a video encoder.

Hereinafter, embodiments of the present invention will be described in detail with reference to the illustrated drawings.

2 is a block diagram showing a video encoder according to the present invention. The video encoder according to the present invention includes a frame forming unit 202, a motion compensation means, an encoding means, a decoding means, a VLC 226, a buffer 228, It is composed of an image separator 230 and an image controller 240.

The frame forming unit 202 is configured to implement the input image signal in units of frames and to provide the image separating unit 230 to be described later.

The image separator 230 stores a frame provided from the frame forming unit 202 as a previous frame and outputs it to the subtractor 204 and the motion predictor 218, which will be described later. The difference signal between the current frames is obtained, and whether the moving picture or the still image of the current frame is determined, and the luminance and color control information of the current frame are analyzed and the image control unit 240 is controlled to be described later in response to the determination result. do.

The motion compensating means includes a subtractor 204, an adder 214, a frame memory 216, a motion predictor 218, and a motion compensator 220, and a previous frame stored in the frame memory 216. It is configured to obtain a motion displacement by the motion prediction between the input frames, and to provide the adder 214 with a predictive frame which is compensated from the previous frame by this motion displacement.

The encoding means is composed of a DCT 206 and a quantization unit 208, and differentially discriminates the difference signal between the current frame and the predicted frame input by control signals provided from the image control unit 240 and the buffer 228, which will be described later. It is configured to be provided to the decoding means and the VLC 226 to be encoded and described later.

The decoding means is composed of an inverse quantization unit 210 and an inverse DCT 212, and decodes the image data differentially encoded by the encoding means by inverse quantization and inverse DCT, and the decoded image data and motion are compensated for. The combined displacements are configured to be stored in the previous frame on the frame memory 216.

The VLC 226 is configured to provide variable length coding of the video data encoded by the encoding means to the buffer 224 described later.

The buffer 224 provides a control signal according to the storage capacity of the variable length coded data provided from the VLC 226 to the quantization unit 208 and the image control unit 240, and converts the variable length coded image data into a decoder. Is configured to be output.

The image controller 240 discriminates based on the difference signal between the current frame and the predicted frame based on the moving picture determination control information and the luminance and color control information for the analyzed current frame provided from the image analyzer 230. It is configured to control the encoding operation of the DCT 206 and the quantization unit 208 to be encoded.

Referring to the present invention configured as described in detail as follows.

First, when the input image signal is provided to the frame forming unit 202 in the form of a bit stream, the frame forming unit 202 implements the image input in the form of a bit screen in units of frames and provides the image separating unit 230 to the image forming unit 230. do.

The image separator 230 stores a frame provided from the frame forming unit 202 as a previous frame, and provides the frames to the subtractor 204 and the motion predictor 218 for differential signal and motion prediction.

In addition, the image separator 230 obtains a difference signal between a next input frame and a previous frame temporarily stored in the image separator 230, where the difference signal between the obtained two frames and a predetermined threshold are obtained. If the difference signal between the two frames is smaller than the preset threshold value, the value is regarded as a still image with small motion and the control information corresponding thereto is provided to the image controller 240.

Accordingly, the image controller 240 performs the same encoding as the conventional DCT coefficients and the quantization coefficients by providing them to the DCT 206 and the quantization unit 208, respectively.

That is, the subtractor 204 obtains the difference signal between the frame output from the image separator 230 and the frame predicted and compensated by the encoding means, encodes the difference signal, and then inverses the quantizer 210 and the VLC 226. ) Will be provided.

Accordingly, the VLC 226 performs variable length encoding on the image data encoded by the DCT 206 and the quantization unit 208 based on the control information of the image control unit 240 using zigzag scanning, and then uses a buffer ( 228).

The buffer 228 outputs the variable length coded data provided from the VLC 226 to the decoder through a transmission channel having a limited frequency band, while quantizing when the capacity of the data of the buffer 224 becomes larger. In addition to providing a control signal that increases the coefficient and lowers the coding rate to the quantization unit 208, the same control signal is also provided to the image control unit 240.

In addition, the inverse quantization unit 210 inversely quantizes the quantized image data provided from the quantization unit 208 and then inverse DCT through the inverse DCT 212 and then provides the adder 214 to the adder 214. ) Is provided with a motion displacement whose motion is compensated for the previous frame.

Therefore, the adder 214 stores the image data obtained by combining the inverse DCT data and the motion displacement as a previous frame in the frame memory 216, and the frame reconstructed therefrom is a motion predictor 218 and a motion compensator ( 220 each.

The motion predictor 218 predicts the motion between the frame reconstructed from the frame memory 216 and the frame provided from the image separator 230, and provides the predicted motion displacement to the motion compensator 220.

The motion compensator 220 switches the predicted frame that compensates for the motion from the reconstructed frame based on the frames provided from the frame memory 216 and the motion displacements provided from the motion predictor 218. 224 to subtracter 204 and adder 214, respectively.

On the contrary, the image separator 230 compares the difference signal obtained from the difference signal between the next input frame and the previous frame temporarily stored in the image separator 230 with a preset threshold. .

As a result of the comparison, when the difference signal between the two frames is larger than the preset threshold, the motion is regarded as a large moving image, and the control information corresponding thereto, that is, the luminance signal representing the outline is more faithfully encoded for the image having a lot of motion. In addition, the color signal is provided to the image controller 240 to provide a control signal that is roughly encoded in consideration of human visual characteristics.

The image controller 240 may provide the DCT coefficients corresponding to the control information provided from the image separator 230 and the quantization adjustment values of the quantization unit to the DCT 206 and the quantization unit 208, respectively.

Accordingly, the DCT 206 and the quantization unit 208 readjust the DCT and the quantization coefficient values according to the adjustment value of the DCT coefficient provided from the image control unit 240 and the quantization adjustment value of the quantization unit.

The subtractor 204 obtains the difference value between the frame stored in the image separator 230 as a previous frame and the predicted frame, and converts the difference signal (the signal according to the moving picture) to the DCT 206 and the quantizer 208. ) Will be provided sequentially.

Accordingly, the DCT 206 encodes the luminance signal representing the adjustment value contour of the DCT coefficient provided from the image controller 240 in more detail, and the color (color) signal is roughly encoded based on the control signal which can be roughly encoded. The difference signal between the two frames by the signal 204 (the signal according to the moving video) is DCT.

In addition, the quantization unit 208 quantizes the coefficient DCTed by the DCT 206 based on the quantization adjustment value provided from the image control unit 240, and then provides the coefficient to the inverse quantization unit 210 and the VLC 226. Done. In other words, the color signal of the fast-moving part limits the color of the high frequency that is not detected by the human visual characteristics and the color that cannot be detected by DCT and quantization.

Accordingly, the VLC 226 variable length encodes the image data quantized by the quantization unit 208 and then outputs it to the decoder through the transmission channel through the buffer 228.

In addition, the inverse quantization unit 210 inversely quantizes the quantized image data (the luminance signal is encoded in detail and the color signal is roughly encoded) provided from the quantization unit 208 and then inversed through the inverse DCT 212. The DCT is then provided to the adder 214.

The adder 214 stores image data obtained by combining the inverse DCT data and the motion displacement as a previous frame in the frame memory 216, and the frame reconstructed therefrom is a motion predictor 218 and a motion compensator 220. Is provided.

The motion predictor 218 predicts the motion between the frame reconstructed from the frame memory 216 and the frame provided from the image separator 230, and provides the predicted motion displacement to the motion compensator 220.

The motion compensator 220 switches the predicted frame that compensates for the motion from the reconstructed frame based on the frames provided from the frame memory 216 and the motion displacements provided from the motion predictor 218. 224 is provided to the subtractor 204 and adder 214 to be used for the next encoding.

As described above, the present invention has an effect of more efficiently compressing image data to be transmitted by discriminating still images and moving images and selectively performing encoding accordingly.

Claims (2)

The discrete cosine transform converts the difference signal between the current frame and the predictive frame into a DCT transform coefficient of the N × N block, quantizes the transformed DCT transform coefficients to a predetermined quantization step size, and DCT of the quantized N × N. In an image encoder for variable-length encoding transform coefficients, a difference between a current frame and a next frame is calculated, and a comparison between the calculated difference and a preset threshold determines whether the current frame is a large motion image or a small motion image. Image separation means for detecting and generating encoding adaptive control information corresponding to the detection result; Means for generating a DCT coefficient adjustment signal for DCT and a quantization coefficient adjustment signal for quantization based on the generated encoding adaptive control information; DCT means, when the DCT coefficient adjustment signal is provided, to DCT the luminance component of the difference signal between the current frame and the predicted frame relatively tightly compared to the color component, and to DCT the rough component relatively roughly to limit the high frequency component; And when the quantization coefficient adjustment signal is provided, quantizes the DCT transform coefficients of the luminance component in a step relatively smaller than the DCT transform coefficients of the color component, and quantizes the DCT transform coefficients of the color component in a relatively large step. Video encoder comprising means. The method of claim 1, wherein the image separation unit detects the current frame as an image having a lot of motion when the calculated difference is greater than the predetermined threshold value, and when the calculated difference is smaller than the predetermined threshold value. The video encoder of claim 1, wherein the current frame is detected as an image with little motion.