KR100285593B1 - Image signal adaptive encoding device - Google Patents

Abstract

The present invention relates to a suitable encoding device for adaptively encoding an image signal comprising a texture signal comprising an object pixel and a background pixel and an outline signal distinguishing the object pixel from the background pixel in the texture signal and the background pixel. Each of the background pixels is represented by luminance data and chromaticity data, and the encoding apparatus includes a padding circuit, a luminance signal extraction circuit, and a frame / field correlation evaluation channel. The padding circuit converts the value of each background pixel into a pixel value derived from the object pixel value using the contour signal according to a predetermined padding method to generate a padded texture signal, and the luminance signal extraction circuit generates a padded texture signal from the padded texture signal. Extracting the luminance data to provide an improved texture signal, wherein each pixel of the improved texture signal has luminance data. In addition, the frame / field correlation evaluation channel evaluates the frame correlation of the frame of the improved texture signal and the field correlation of the upper field and the lower field of the texture signal based on the contour signal and the improved texture signal. Based on the frame encoding mode signal or the field encoding mode signal is adaptively generated.

Description

Image signal adaptive encoding device

The present invention relates to an image signal adaptive encoding apparatus for adaptively encoding an image, and more particularly, to an image signal adaptive suitable for adaptively encoding an image signal based on a frame / field correlation of the image signal. It relates to an encoding device.

Generally in digital television systems such as video telephony, teleconference and high definition television systems, the video line signal of a video frame signal contains digital data called pixel values, so that a significant amount of each video frame signal is required to define each video frame signal. Digital data is required.

However, the frequency bandwidth available in conventional transmission channels is limited, especially in low bit-rate video signal encoders such as video telephony or teleconferencing systems. The amount should be reduced.

As is well known, one of the techniques for encoding video signals used in low rate coding systems is the so-called ″ object-oriented analysis-sysnthesis coding technique ″. In, the input video signal is divided into a plurality of objects, and variables for defining each object's motion, contour, and pixel data are processed through different coding channels.

One embodiment of the object-oriented analysis and synthesis coding scheme is a so-called video expert group 4 (MPEG-4), where MPEG-4 is content-based bidirectionality in applications such as low bit rate communication, interactive multimedia and monitors. It is intended to propose an audio-video encoding standard for satisfying content-based interactivity, improved coding efficiency and / or versatility.

According to MPEG-4, an input image is divided into a plurality of video object planes (VOPs), which correspond to entities of bit streams that are accessible and manipulated by a user. Since the VOP is represented by a bounding rectangle having a minimum size among squares having an integer multiple of 16 pixels surrounding the object, the encoder may process the input image in units of VOP.

The VOP includes a shape signal and a texture signal composed of luminance data and chromaticity data, wherein the shape information is represented, for example, as a binary mask. In the binary mask, for example, a binary value such as 0 is used to represent a pixel outside the object in the VOP, i.e. a background pixel, and another binary value such as 255 is used to represent the object in the VOP. It is used to represent a pixel present inside, that is, an object pixel. Prior to encoding an image signal, for example a digital image signal having a VOP or an object, it is preferable to pad the texture signal of the image signal. The reason is to prevent the coding efficiency from being degraded due to the high frequency pixel data in the background region of the image signal or in the region located outside the object. Thus, each background pixel value in the texture signal of the image signal is generally padded with pixel values derived from object pixel values using conventional padding methods.

For example, in the conventional average value padding method, each pixel in the background of an image signal such as, for example, a VOP is padded with an average value of all pixel values in the object of the VOP. According to the conventional repetitive padding method, each pixel value in the background of the VOP, that is, the background pixel value, is padded with a pixel value derived from the boundary pixel value of the VOP (see MPEG-4 Video Verification Model Version 7.0, International Organization for Standardization, Coding of Moving Pictures And Associated Audio Information, ISO / IEC JTC1 / SC29 / WG11 MPEG97 / N1642, Bristol, April 1997, pp 40-41). After padding for the texture signal is completed, the padded texture signal is encoded using a transform encoding technique with or without a motion estimation (ME) technique and a motion compensation (MC) technique.

On the other hand, in terms of scanning form, the video sequence of the image signal is divided into two types, namely, a forward scan video sequence and a perforated scan video sequence. In a progressive scan video sequence, the frames in the sequence are captured and processed line by line in order from top to bottom of the frame. A conventionally scanned video frame consists of two fields: an upper-field (or even-field) made of even-numbered lines of the frame and a lower-field (or odd-field) made of odd-numbered lines of the frame. The capture and processing of these two fields is first done from top to bottom for the top-field and then captured and processed in the same way for the bottom-field.

Conventional researches on image signal coding can be classified into three categories, namely, a frame encoding process, a field encoding process, and an adaptive encoding process using both a frame encoding process and a field encoding process.

In the frame encoding process, a video sequence is basically encoded frame by frame, where the upper-field and the lower-field of the frame are combined in parallel (where the frame is treated as traversal). In the frame encoding process, each frame is usually divided into pixel blocks and then encoded by a transform coding method such as a discrete cosine transform (DCT) encoding method. ″ Scene Adative Coder ″ by Chen and Pratt, IEEE Transaction on Communication, COM-32, No. 3, pp 225-232 (March 1984).

In the field encoding process, the video sequence is first divided into two sequences, namely, the upper-field and the lower-field, and then the upper-field and the lower-field are encoded in the same manner as the frame encoding in the same manner as the frame encoding.

In this field, the frame encoding process is effective when the fixed region in the image signal is encoded using a high spatial correlation therein, and the field encoding process is applicable in encoding a region corresponding to a moving object in the image signal. It is known that coding can be performed more efficiently than frame coding when there is a higher correlation in each field in the region. Accordingly, in the adaptive encoding process, the image signal is encoded in a frame unit if the correlation of the frame of the image signal is higher than the correlation of the upper-field and the lower-field of the frame.

In recent years, the adaptive coding process has attracted much attention because of its high efficiency and flexibility. For example, US Pat. No. 5,347,308 to Lucas et al. Discloses a representative conventional apparatus and method for adaptively encoding image signals. 1 illustrates an apparatus for adaptively encoding an image signal disclosed in US Pat. No. 5,347,308 to Lucas et al. In the apparatus of Lucas et al., The frame 1 of the input image signal is divided into a plurality of pixel data blocks by a block division process (2).

Then, the block, i.e. each block, enters an inter-field difference detection process 3 for examining the difference between the pixel data between two fields in the block. It should be noted that the inter-field difference detection process 3 can be regarded as a process of calculating (or evaluating) a kind of frame / field correlation for each block.

An example of the inter-field difference detection process 3 is as follows. That is, contiguous with the MSE between successive even-numbered line pairs in the block equal to the first mean square error (MSE) between the even-numbered line in the block and the line pair including the odd-numbered line adjacent thereto. After calculating a second MSE that sums the MSEs between odd-numbered line pairs, and then the ratio of the first MSE to the second MSE is greater than a predetermined threshold, block 4, e.g., block 4 shown in FIG. Otherwise, the block is encoded by the frame encoding process (6).

The coded data for each block from the field encoding process 5 and the frame encoding process 6 is transmitted as channel information, respectively. The result data 7 obtained by the inter-field difference detection process 3 for each block is encoded and transmitted as side information or additional information for each block.

However, Lucas et al. Perform the inter-field difference detection process (4) for all blocks in the image signal and encode and transmit additional information in block units for all blocks in the image signal. Generate transmission data.

Moreover, it is known in the art that the contour of an object in the image signal is very important for estimating the motion of the object in encoding the image signal. However, a device by Lucas et al. Or any other conventional device and / or method may be used to convert an image signal having an object, for example, a contour signal included in the image signal in the calculation (or evaluation) of the frame / field correlation of the VOP. I didn't use it. Thus, conventional apparatus and / or methods have failed to reduce the amount of side information for the image signal while making the process of calculating the frame / field correlation (or evaluation) for the image signal simple and effective. Therefore, the conventional apparatus and method have a limit in increasing the coding efficiency of an image signal.

Accordingly, an object of the present invention is to simplify and effectively calculate (or evaluate) the frame / field correlation of an image signal based on the frame / field correlation of the image signal calculated using the contour signal of the image signal. The present invention provides an image signal adaptive encoding apparatus capable of increasing encoding efficiency by reducing an amount of additional information about an image signal.

In order to achieve the above object, the present invention provides an image for adaptively encoding an image signal including a texture signal comprising an object pixel in an object and a background pixel in a background and a contour signal that distinguishes the object pixel from the background pixel. In the signal adaptive encoding apparatus, each of the object pixel and the background pixel is represented by luminance data and chromaticity data. The apparatus uses the contour signal according to a predetermined padding method to determine the value of each background pixel. Padding means for converting the pixel value derived from the object pixel value into a padded texture signal; Brightness data extraction means for extracting luminance data from the padded texture signal to provide an improved texture signal, wherein each pixel of the enhanced texture signal has luminance data; And based on the contour signal and the improved texture signal, the frame correlation of the frame of the enhanced texture signal and the field correlation of the upper and lower fields of the texture signal are evaluated, and then the frame correlation is determined by the field. If the correlation is higher than that, the image signal is determined to be encoded in units of frames to generate a frame encoding mode signal. If the frame correlation is not higher than the field correlation, the frame is determined to be encoded in units of fields. And a frame / field correlation evaluation means for generating a mode signal and calculating the frame correlation and the field correlation according to a predetermined rule.

1 is a block diagram of a conventional image signal adaptive encoding apparatus,

2 is a block diagram of an image signal adaptive encoding apparatus according to an embodiment of the present invention;

3 shows an improved texture signal comprising a contour block;

4 is a view illustrating a contour block formed by combining an upper-field block and a lower field block;

FIG. 5 is a detailed block diagram of the block division circuit shown in FIG. 2 in accordance with one preferred embodiment of the present invention. FIG.

<Description of the code | symbol about the principal part of drawing>

210: padding circuit 220: luminance data extraction circuit

231: contour block detection circuit 232: correlation calculation circuit

233: encoding mode determination circuit 240: block division circuit

241: encoding path selection circuit 245: frame-block division circuit

247: field-block division circuit 251: frame encoding circuit

252: field encoding circuit 253: data formatting circuit

260: multiplexer (MUX) 400: contour block

410: upper-field block 420: lower-field block

2, a block diagram of an image signal adaptive encoding apparatus 200 according to a preferred embodiment of the present invention is shown. FIG. 2 shows a detailed block diagram of the fourth decision unit shown in FIG. 1. The image signal includes a texture signal having an object pixel in the object and a background pixel in the background. The image signal also includes an outline signal that distinguishes the object pixel from the background pixel. Each of the object pixel and the background pixel in the texture signal is represented by luminance data and chrominance data, and a representative image signal is a video object plane (VOP). And the image signal may also be a digital image frame containing an object.

The contour signal has mask data having a first binary value and a second binary value, where a first binary value, for example 255, is used to designate an object pixel, for example, a second, such as zero. Binary values are used to specify background pixels. Therefore, the mask data included in the contour signal may be used to distinguish the object and the background in the texture signal.

The encoding apparatus 200 may include a padding circuit 210, a luminance data extraction circuit 220, a frame / field correlation evaluation channel 230, a block division circuit 240, a texture signal encoding channel 250, and a multiplexer (MUX). 260. In addition, the frame / field correlation evaluation channel 230 includes an outline block detection circuit 231, a correlation calculation circuit 232, and an encoding mode determination circuit 233, and the texture signal encoding channel 250 includes frame encoding. A circuit 251, a field encoding circuit 252, and a data formatting circuit 253.

For adaptive encoding of the image signal, first, the contour signal of the image signal is input to the contour block detection circuit 231, the padding circuit 210 and the MUX 260 via the line L10. The texture signal of the image signal is provided to the padding circuit 210 through the line L20, and the padding circuit 210 performs padding on the texture signal.

In more detail, the padding circuit 210 uses the contour signal input there through the line L10 according to a predetermined padding technique, for example, a conventional padding technique or a repetitive padding technique. The padded texture signal is generated on the line L30 by converting the pixel value derived from the object pixel value to the pixel value derived from the object pixel value and transmitted to the luminance data extraction circuit 220 and the block division circuit 240 through the line L30, respectively.

The luminance data extraction circuit 220 extracts the luminance data from the padded texture signal and provides the improved texture signal to the outline block detection circuit 231 through the line L40. Each pixel of the improved texture signal includes luminance data. Has

The frame / field correlation evaluation channel 230 evaluates the frame correlation of the frame of the enhanced texture signal and the field correlation of the upper field and the lower field of the texture signal based on the contour signal and the improved texture signal. If the correlation is higher than the field correlation, the image signal is determined to be encoded frame by frame, and if the correlation is not higher than the field correlation, the frame encoding mode signal is generated. In generating the mode signal, the frame correlation and the field correlation are calculated according to a predetermined rule. Examples of predetermined rules are described next.

More specifically, the contour block detection circuit 231 in the frame / field correlation evaluation channel 230 uses the same MxN using the contour signal input there through line L10 and the improved texture signal input there through line L40. A plurality of contour blocks of pixel size are detected and fed to the correlation calculation circuit 232, where M and N are each predetermined positive integers. Here, it should be noted that the frame of each contour block is a block having at least one object pixel and at least one background pixel, and the frame of the contour block is made by combining an upper field with even lines and a lower field with odd lines. There is a need.

3 shows an improved texture signal 300 including contour block 310, wherein the hatched and unhatched areas of the improved texture signal 300 represent objects and a background, respectively. 4 illustrates a 16x16 pixel sized outline block 400 formed by combining a 16x8 pixel sized top-field block 410 and a 16x8 pixel sized bottom field block 420. Contour block 400 consists of 16 horizontal lines numbered from 0 to 15 as shown in FIG. The hatched and unhatched areas in the contour block 400, the upper field block 410 and the lower field block 420 represent the even and odd lines, respectively.

The correlation calculation circuit 232 calculates a block-frame correlation value (hereinafter referred to as BRCV) and a block-field correlation value (hereinafter referred to as BDCV) for the contour block input therefrom from the contour block detection circuit 231. To generate a plurality of BRCV values and a plurality of BDCV values for all of the plurality of contour blocks, respectively, on line L45, where BRCV is a correlation value calculated according to a predetermined rule for the frame of the contour block and BDCV is Correlation values calculated according to predetermined rules for the top-field block and the bottom-field block of the contour block.

According to one preferred embodiment of the present invention, the BRCV value is a block-frame difference (hereinafter referred to as BRD) for each contour block and the BRDV value is a block-field difference (hereinafter referred to as BDD) for each contour block. . BRD is the sum of the absolute values of the plurality of first differences, where each first difference is the error between the even line of each contour block and the line pair comprising adjacent odd lines; The BDD is the sum of the absolute values of the plurality of second differences and the absolute values of the plurality of third differences, where each second difference is an error between successive even line pairs of each contour block and each third difference is The error between successive odd line pairs of each contour block.

According to another preferred embodiment of the present invention, the BRCV value is the block-frame error squared for each contour block (hereinafter referred to as BRS) and the BRDV value is the block-field error squared for each contour block (hereinafter referred to as BDS). ). The BRS is the sum of a plurality of first error squares, where each first error squared is the square of the error between the even line of each contour block and the line pair comprising adjacent odd lines; The BDS is the sum of the plurality of second error squares and the plurality of third error squares, where each second error square is the square of the error between successive even line pairs of each contour block and each third error square is The square of the error between successive odd-numbered line pairs of each contour block.

Hereinafter, the correlation calculation process performed by the correlation calculation circuit 232 will be described in detail with reference to FIG. 4. In this case, for convenience of description, it is assumed that the size of the contour block is 16x16 pixel size, that is, M and N are each 16.

According to a preferred embodiment of the present invention, BRD and BDD are calculated by the following equations (1) and (2), respectively.

In the above equation, P _h and _v are luminance values of pixels located at the intersection of the h-th horizontal line and the v-th vertical line in each contour block, and h and v are numbers ranging from 0 to 15, respectively.

According to another preferred embodiment of the present invention, BRS and BDS are calculated by the following equations (3) and (4), respectively.

In the above equation, P _h and _v are luminance values of pixels located at the intersection of the h-th horizontal line and the v-th vertical line in each contour block, and h and v are numbers ranging from 0 to 15, respectively. In Equations 1 to 4, the horizontal line number h and the vertical line number v are numbered in increasing order from top to bottom and from left to right in the contour block, respectively.

The encoding mode determination circuit 233 performs encoding mode determination based on the plurality of BRCV values and the plurality of BDCV values to encode the image signal on a field-by-field basis. Generate one of the commanded field coding mode signals.

In more detail, according to a preferred embodiment of the present invention, the encoding mode determination circuit 233 first counts a first number representing the number of first contour blocks and a second number representing the number of second contour blocks. Each first contour block is an contour block having a corresponding BRD equal to or less than a corresponding BDD and each second contour block is a contour block having a corresponding BRD larger than the corresponding BDD. The encoding mode determining circuit 233 then generates a frame encoding mode signal on the line L50 if the first number is equal to or greater than the second number, and generates a field encoding mode signal if the first number is less than the second number. Create on line L50.

According to another preferred embodiment of the present invention, the encoding mode determination circuit 233 generates a first sum by adding corresponding BRSs for all contour blocks and generating a second sum by adding corresponding BDSs for all contour blocks. . Then, the encoding mode determination circuit 233 compares the first sum and the second sum to generate the frame encoding mode signal if the ratio of the second sum of the first sum is equal to or less than a predetermined threshold, for example, 0.8. A field coding mode signal is generated on line L50 if it is generated on line L50 and the ratio for the second sum of the first sum is greater than a predetermined threshold. As a result, one of the frame encoding mode signal and the field encoding mode signal is supplied to the data formatting circuit 253 via the line L50.

The block dividing circuit 240 divides the padded texture signal input therefrom from the padding circuit 210 through the line L30 in response to the frame encoding mode signal from the encoding mode determining circuit 233 to obtain a plurality of identical KxL pixel sizes. The frame-block is provided to the frame encoding circuit 251 through the line L60. And the block division circuit 240 transfers the padded texture signal input therefrom from the padding circuit 210 to the top-field and the bottom-field in response to the field coding mode signal from the coding mode determination circuit 233. After separating, the upper-field is divided into upper field-blocks of the same KxL pixel size, while the lower-field is divided into lower field-blocks of the same KxL pixel size and the upper field- of multiple identical KxL pixel sizes. A lower field-block having a block and a plurality of identical KxL pixel sizes is supplied to the field encoding circuit 252 through a line L70. In the above description, each KxL pixel sized frame-block supplied to the frame encoding circuit 251 through the line L60 has only an upper field-block and an odd numbered line having a Kx (L / 2) pixel size having only even-numbered lines. A lower field-block having a Kx (L / 2) pixel size and a lower field-block having a KxL pixel size and a KxL pixel size supplied to the field encoding circuit 252 through a line L70. Has only even lines and only odd lines, respectively.

FIG. 5 shows a detailed block diagram of the block division circuit 240 shown in FIG. 2 in accordance with one preferred embodiment of the present invention. Referring to FIG. 5, a block division process performed by the block division circuit 240 will be described in more detail below. The block division circuit 240 includes an encoding path selection circuit 241, a frame-block division circuit 245, and a field-block division circuit 247.

The encoding path selection circuit 241 selects an encoding path for each frame in response to the frame encoding mode signal from the encoding mode determination circuit 233, and selects a padded texture signal input therefrom from the padding circuit 210 through the line L30. Provided as a frame to frame-block dividing circuit 245 via line L41. The encoding path selection circuit 241 selects the encoding path for each field in response to the field encoding mode signal from the encoding mode determination circuit 233, and outputs a padded texture signal input therefrom from the padding circuit 210 to the upper field. It is divided into the lower field and provided to the field-block division circuit 247 through the line L43.

The frame-block dividing circuit 245 divides the frame input therein into a plurality of frame-blocks having the same KxL pixel size and provides it to the frame encoding circuit 251 through the line L60.

The field-block splitting circuit 247 divides the top-field from the coding path selection circuit 241 into a plurality of top-field blocks of the same KxL pixel size and provides it to the field coding circuit 252 through the line L70. At the same time, the field-block dividing circuit 247 divides the sub-field input there through the line L43 into a plurality of sub-field blocks of the same KxL pixel size and provides it to the field encoding circuit 252 via the line L70.

When the frame encoding mode signal is input there through the line L50, the texture signal encoding channel 250 encodes the frame-block input thereto through the line L60 frame by frame to provide the encoded texture signal to the MUX 260 through the line L95. do. When the field encoding mode signal is input there through the line L50, the texture signal encoding channel 250 encodes the upper-field block and the lower-field block input thereto through the line L70 to MUX the encoded texture signal through the line L95. Provided at 260.

In more detail, the frame coding circuit 251 in the texture signal coding channel 250 includes a Discrete Cosine Transform (DCT) coding method and a quantization method of a frame block, and motion estimation (ME). ) And a frame-block encoded by using a conventional frame encoding technique including or not including a motion compensation (MC) scheme is provided to the data formatting circuit 253 through a line L80.

The field encoding circuit 252 encodes the upper-field block and the lower-field block by encoding using a conventional field encoding technique including a DCT encoding method and a quantization method and including or not including ME and MC techniques. Provide the field block and the encoded lower-field block to the data formatting circuit 253 via line L80. The above-described DCT coding technique is performed in units of blocks, and the block size is typically 8x8 pixel size, where K and L are 8 respectively.

When the frame encoding mode signal is input thereto through the line L50, the data formatting circuit 253 combines the frame encoding mode signal and the encoded frame-block and provides the encoded texture signal to the MUX 260 through the line L95. When the field encoding mode signal is input there through the line L50, the data formatting circuit 253 combines the field encoding mode signal with the encoded upper-field block and the encoded lower-field block to encode the texture signal through the line L95. Provided to the MUX 260. It is to be noted that the frame encoding mode signal and the field encoding mode signal are some kind of additional information about the encoded texture signal, respectively.

MUX 260 multiplexes the contour signal input there through line L10 and the encoded texture signal input there through line L95 and provides it to the transmitter (not shown) for transmission as an encoded image signal.

As described above, according to the present invention, unlike the prior art in which the additional information on the encoded texture signal is added to all of the plurality of blocks made by dividing the texture signal, the additional information on the encoded texture signal is added to the frame of the texture. Alternatively, additional information may be added in units of fields of a frame of the texture, thereby greatly reducing the amount of additional information, thereby improving encoding efficiency.

While specific embodiments of the invention have been described above, those skilled in the art can make various changes without departing from the scope of the claims described herein.

As described above, according to the present invention, by adaptively encoding the image signal based on the frame / field correlation of the image signal calculated using the contour signal of the image signal, the frame / field correlation of the image signal is It is possible to increase the coding efficiency by reducing the amount of additional information while making the evaluation process simple and effective.

Claims (8)

An image signal adaptive encoding apparatus for adaptively encoding an image signal including a texture signal including an object pixel in an object and a background pixel in a background, and a contour signal distinguishing the object pixel from the background pixel And each of the background pixels is represented by luminance data and chromaticity data. Padding means for generating a padded texture signal by converting a value of each background pixel into a pixel value derived from the object pixel value using the contour signal according to a predetermined padding method; Brightness data extraction means for extracting luminance data from the padded texture signal to provide an improved texture signal, wherein each pixel of the enhanced texture signal has luminance data; And Based on the contour signal and the improved texture signal, the frame correlation of the frame of the improved texture signal and the field correlation of the upper and lower fields of the texture signal are evaluated, and then the frame correlation is the field correlation. If it is higher than this, it is determined to encode the image signal in units of frames to generate a frame encoding mode signal. If the frame correlation is not higher than the field correlation, it is determined that the image signals are encoded in units of fields. Frame / field correlation evaluation means for generating a signal and calculating the frame correlation and the field correlation according to a predetermined rule. The method of claim 1, wherein the frame / field correlation evaluation means is: The contour signal and the improved texture signal are used to detect a plurality of contour blocks having the same MxN pixel size to supply the frame of the contour block, wherein each contour block includes one or more object pixels and one or more background pixels. A frame of the contour block is formed by combining an upper field having an even-numbered line and a lower field having an odd-numbered line, wherein M and N are contour block detecting means each having a predetermined positive integer; Calculating a block-frame correlation value (BRCV) and a block-field correlation value (BDCV) for each contour block to generate a plurality of BRCV values and a plurality of BDCV values for all of the plurality of contour blocks, The BRCV is a correlation value calculated according to a predetermined rule for the frame of each contour block, and the BDCV is calculated according to a predetermined rule for the upper-field block and the lower-field block of each contour block. Correlation value calculation means which is a calculated correlation value; And A frame encoding mode signal for determining the encoding mode of the image signal based on the plurality of BRCV values and the plurality of BDCV values and for encoding the image signal for each frame and for encoding the image signal for each field. And encoding mode determining means for generating one of the field encoding mode signals. 3. The method of claim 2, wherein the BRCV value is a block-frame difference (BRD) for each contour block and the BRDV value is a block-field difference (BDD) for each contour block, wherein the BRD is a plurality of A sum of absolute first differences, wherein each first difference is an error between an even line of each contour block and a line pair comprising adjacent odd lines; The BDD is the sum of the plurality of second difference absolute values and the plurality of third difference absolute values, where each second difference is an error between successive even-numbered line pairs of each contour block and each third difference is And an error between successive odd-numbered line pairs of each contour block. The method of claim 3, wherein the encoding mode determining means counts a first number representing the number of first contour blocks and a second number representing the number of second contour blocks, and then the first number is equal to the second number. Is greater than or equal to, generating a frame encoding mode signal, and if the first number is less than the second number, generating a field encoding mode signal, wherein each first contour block has an outline having a corresponding BRD equal to or less than a corresponding BDD. And each of the second contour blocks is a contour block having a corresponding BRD larger than the corresponding BDD. 5. The apparatus of claim 4, wherein the encoding apparatus divides the padded texture signal in response to the frame encoding mode signal to provide a plurality of frame-blocks having the same KxL pixel size, and the padding in response to the field encoding mode signal. The separated texture signal into an upper-field and a lower-field, and then divide the upper-field to provide a plurality of upper field-blocks having the same size of KxL pixels, and simultaneously split the lower-fields into a plurality of identical KxL pixels. And image segmentation means for supplying a lower field-block of size. The method of claim 5 wherein said dividing means is: A coding path for each frame is selected in response to the frame coding mode signal to provide the padded texture signal as a frame, and a coding path for each field is selected in response to the field coding mode signal to display the padded texture signal in an upper field and a lower part. Encoding path selecting means for dividing the information into fields; The frame is divided into a plurality of frame blocks of the same size of KxL pixels, and each of the frame blocks of size of KxL pixels includes an upper field block having a Kx (L / 2) pixel size having only even-numbered lines. Frame-block dividing means made by combining lower field-blocks having a Kx (L / 2) pixel size having only odd-numbered lines; And The upper field is divided into a plurality of upper field-blocks having the same KxL pixel size, and the lower field is divided into a plurality of lower field-blocks having the same KxL pixel size, and the upper field-block is larger than the KxL pixel size. And has only even-numbered lines and the lower field-block of the KxL pixel size includes field-block dividing means having only odd-numbered lines. 3. The method of claim 2, wherein the BRCV value is block-frame error squared (BRS) for each contour block, the BRDV value is block-field error squared (BDS) for each contour block, and the BRS Is a sum of a plurality of first error squares, and each first error squared is an error squared between a pair of lines including an even-numbered line and an adjacent odd-numbered line of each contour block; The BDS is the sum of the plurality of second error squares and the plurality of third error squares, each second error square is the error square between successive even line pairs of each contour block, each third error Square is an error square between successive odd-numbered line pairs of each contour block. 8. The method of claim 7, wherein the encoding mode determining means generates a first sum by adding corresponding BRSs for all the contour blocks, and generating a second sum by adding corresponding BDSs for all the contour blocks, and then generating the first sum. Comparing the second sum with the second sum to generate a frame encoding mode signal if the ratio for the second sum is equal to or less than a predetermined threshold and the ratio for the second sum of the first sum is predetermined. And a field encoding mode signal is generated when the threshold value is larger than the threshold value.