KR0138291B1 - Video encoding of digital video cassette recorder - Google Patents

Abstract

The present invention relates to a video encoder of a digital video cassette recorder, comprising: a block formatter for formatting an input video frame data into a predetermined number of macroblocks; A general data compression encoder for encoding the macro block according to a predetermined compression rate to generate general playback data; A trick data compression encoder configured to form a trick block by sampling the macro block at a predetermined ratio, and compressing and encoding a predetermined amount of bits to generate trick reproduction data; And a multiplexer for alternately outputting the general reproduction data and the trick reproduction data at regular intervals.

According to the present invention, there is an advantage that the image quality for trick playback can be improved by fixing the data in several block units and recording the data for trick playback at regular intervals.

Description

Video Encoder of Digital Video Cassette Recorder

1 is a block diagram of an image encoder according to the present invention.

2 is an example of a block reformatting form related to FIG.

3 is a data recording form of a tape according to the present invention.

4 to 7 show examples of trick play data processing according to the present invention.

* Description of the symbols for the main parts of the drawings *

11... Block formatter 12... Discrete Cosine Transformation Unit

13... Quantization unit 14. Variable length coding

15... Block reformatter 16... Discrete Cosine Transformation Unit

17... Trick playback data processor 18. Multiplexer

19... Error Correction Code

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video encoder of a digital video cassette recorder (hereinafter referred to as DVCR), and more particularly to a DVCR for dividing normal playback data and trick playback data on a tape and alternately recording them. It relates to a video encoder.

The algorithm for encoding data for trick reproduction of DVCR has a limitation that each segment must be decoded independently, and the amount of bits to be coded per segment for trick reproduction must be fixed. Accordingly, in DVCR, a special tape format is required, and a bit arrangement method that can be controlled while going forward with respect to a fixed output bit amount by each segment unit is required.

In general, the bit allocation method uses a feedback control by a rate buffer in a communication and a random access recording system, but cannot be applied to a recording medium which must be accessed sequentially such as a tape.

In the DVCR, when only normal playback data is encoded by video compression encoding without recording data for normal playback and trick playback, the trick playback data is not processed separately, resulting in deterioration in image quality during trick playback.

Accordingly, it is an object of the present invention to provide a video encoder of a DVCR capable of improving image quality during trick playback by recording general playback data and trick playback data at regular intervals on a tape.

The present invention for achieving the above object,

In the video encoder of the digital video cassette recorder for video compression encoding the input video frame data,

A block formatter for formatting the frame data into a predetermined number of macroblocks;

A general data compression encoder for encoding the macro block according to a predetermined compression rate to generate general playback data;

A trick data compression encoder configured to form a trick block by sampling the macro block at a predetermined ratio, and compressing and encoding a predetermined amount of bits to generate trick reproduction data; And

And a multiplexer for alternately outputting the general reproduction data and the trick reproduction data at regular intervals.

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

FIG. 1 is a block diagram showing the structure of a DVCR video encoder according to the present invention, in which a data is encoded by encoding and encoding a data for general reproduction and a general data encoding unit for generating general reproduction data. A trick data coding unit for generating trick playback data by compressing and encoding data for playback is provided in parallel, and the multiplexer 18 controls the bit amount so that normal playback data and trick playback data can be recorded at regular intervals. After the encoding process for error correction is performed in the error correcting encoder 19, it is recorded on a tape.

The general data encoder comprises a block formatter 11, a first discrete cosine transformer 12, a quantizer 13, and a variable length coder 14 It is composed of

The block formatter 11 formats the input frame data in units of macro blocks of Y, Cr, and Cb signals. Each macro block consists of four 8x8 pixel DCT blocks for the Y signal and two 8x8 pixel DCT blocks for the Cr and Cb signals.

The first discrete cosine transforming unit 12 converts an image block of a constant size, that is, data in the time domain, for example, an 8x8 pixel DCT block into 64 discrete cosine transform coefficients in the spatial frequency domain, and simultaneously performs energy compression. do.

The quantization unit 13 quantizes the video signal energy compressed by the discrete cosine transforming unit at a predetermined quantization interval, and the variable length encoding unit 14 generates encoded data having a variable length in accordance with the amount of information of the data so as to minimize the overall bit. Encode by number.

Compression encoding of general data is performed in the same way as conventional encoding. The input frame data is formatted in macroblock units, the energy is concentrated in the discrete cosine transform unit, the quantization unit is quantized, and the variable length encoding unit converts the data for general reproduction. Is generated and output to the multiplexer 18. However, the compression rate of the normal playback data is controlled in accordance with the bit amount of the trick playback data to be recorded on the tape.

The trick data coding unit includes a block reformatter 15, a second discrete cosine transforming unit 16, and a trick reproduction data processing unit 17.

The block reformatter 15 performs block reformatting for compressing and encoding the blocks by trick reproduction data by sampling the blocks at a predetermined rate with respect to some macroblocks formatted by the block formatter 911.

FIG. 2 shows a reformatting form in the case of fixing trick playback data of two macroblocks. The Y signal is divided into four 8xpixel DCT blocks for a 16x16 pixel macroblock, sampled 4x4 pixels for each DCT block, and reconstructed into a new 8x8 pixel trick block, In the same way, a block containing the sum of Cr and Cb signals is reconstructed into one 8x8 pixel trick block.

The second discrete cosine transform unit 16 converts the trick image block, which is data in the time domain, into 64 discrete cosine transform coefficients, which are data in the spatial frequency domain, and simultaneously performs energy compression.

The trick replay data processing unit 17 fixes the discrete cosine transformed data to encoded data of a constant bit amount and outputs it to the multiplexer 18. The output bit amount of the trick play data can be adjusted according to the image quality selection for normal play and trick play.

3 shows a form in which data is recorded on a tape, in which normal playback data 31 and trick playback data 32 are recorded alternately at regular intervals. During trick play, the head moves in the direction 33 in which the trick play data 32 is recorded, and performs trick play using the recorded trick play data 32.

4 to 7 illustrate an embodiment of data processing of the trick data encoder and the relation between the playback data accordingly. In each figure, (a) shows the source input data, (b) the discrete cosine transformed data for the input data, and (c) the inverse DCT for the data of (b). Each playback data is shown.

4 shows a discrete cosine transform 4b for an 8x8 pixel trick block 4a composed of four distinct parts, forming a block with data of slightly separated portions, and the energy is generated near one row and one column. Are concentrated. After encoding and recording all of them, the inverse discrete cosine transform 4c can reproduce the original input values as they are. However, since many bits are recorded on the tape, the compression efficiency of general playback data is reduced.

In FIG. 5, a block 5a is formed from data of an adjacent block, the discrete cosine transform 5c is formed, all the encoded data are encoded and recorded, and then the inverse discrete cosine transform 5c can reproduce the originally inputted values.

6 shows a discrete cosine transform 6b by forming a block 6a with data of adjacent blocks, writes only a DC value (i.e., 230) onto a tape, and inversely discrete cosine transform 6c to obtain the pixel average value of the trick block. It is restored uniformly. Although the trick reproduction efficiency is lowered, it is possible to produce a close image with few bits, and does not greatly affect the compression efficiency of general reproduction data.

7 shows a discrete cosine transform 7b by forming a block 7a from data of adjacent blocks, and a total of 3, such as a DC value and two AC values (distribution level [0] [1], [1] [0]). Data (i.e., 230, -9, -18) is recorded on the tape, and then the inverse discrete cosine transform (7c) can obtain data close to the original input data, and can make an image almost close.

Therefore, a lot of data is recorded on the tape for good screen composition according to the selection of the quality of the trick play, and a small amount of data is recorded on the tape if excellent image quality is not required. The amount of data to be recorded can be selected as appropriate.

As described above, according to the present invention, a general data processing unit for encoding data for general reproduction and a trick data processing unit for encoding data for trick reproduction by fixing the data in several block units are provided in parallel. By fixing the data in units and recording the data for trick playback at regular intervals, there is an advantage in that image quality can be improved during trick playback.

Claims (10)

In a video encoder of a digital video cassette recorder which formats video frame data into a predetermined number of macroblocks, compresses and encodes the macroblocks at a predetermined compression rate, and records data for general playback on a tape. A block reformer for sampling the macroblock at a predetermined ratio and formatting the trick block into trick blocks in a predetermined block unit; A discrete cosine transform unit converting the trick block into a discrete cosine transform coefficient to perform energy compression; And And a trick-play data processing unit for generating trick-play data by encoding the discrete cosine transformed data into a predetermined bit amount. The method of claim 1, wherein the predetermined ratio of the block reformatter is And a 4x4 pixel sample from each DCT block of the macroblock. The method of claim 1, wherein the trick block of the block reformatter And an 8x8 pixel block sampled from the macroblock. The trick playback data of claim 1, wherein the trick playback data processing unit comprises: And an image encoder of the digital video cassette recorder characterized by encoding all the discrete cosine transformed data. The trick playback data of claim 1, wherein the trick playback data processing unit comprises: And a DC value of the discrete cosine transformed data. The trick playback data of claim 1, wherein the trick playback data processing unit comprises: And an image encoder of the digital video cassette recorder characterized by encoding three data such as a DC value among the discrete cosine transformed data and two AC values of arrays [0] [1] and [1] [0]. In the video encoder of the digital video cassette recorder for video compression encoding the input video frame data, A block formatter for formatting the frame data into a predetermined number of macroblocks; A general data compression encoder for encoding the macro block according to a predetermined compression rate to generate general playback data; A trick data compression coder configured to form a trick block by sampling the macro block at a predetermined ratio, and compressing and encoding a predetermined amount of bits to generate trick reproduction data; And And a multiplexer for alternately outputting the general playback data and the trick playback data at regular intervals. The method of claim 7, wherein the predetermined compression rate of the general data compression coder And the video encoder according to the predetermined bit amount of the trick play data. The method of claim 7, wherein the general data compression coder A first discrete cosine transforming unit for converting the macro block into a discrete cosine transforming coefficient to perform energy compression; A quantization unit for quantizing the data compressed by the first discrete cosine transform unit at a predetermined quantization interval; And a variable length encoder which encodes the quantized data to a variable length according to the information amount of the data to generate general playback data. The method of claim 7, wherein the trick data compression coding unit A block reformer for sampling the macroblock at a predetermined ratio and formatting the trick block into trick blocks in a predetermined block unit; A second discrete cosine transforming unit performing energy compression by converting the trick block into a discrete cosine transforming coefficient; And And a trick-play data processing unit for generating trick-play data by encoding the energy-compressed data by the discrete cosine transforming unit by a predetermined amount of bits.