KR100883603B1 - Method and apparatus for decoding video signal using reference picture - Google Patents

Abstract

In a method of decoding a video signal, at least a portion of the current image of the current layer is predicted based on at least a portion of the reference image, offset information and dimension information. The offset information indicates a position offset between at least one boundary pixel of the reference image and at least one boundary pixel of the current image, and the dimension information indicates at least one dimension of the current image.

Description

METHOD AND APPARATUS FOR DECODING VIDEO SIGNAL USING REFERENCE PICTURES}

The present invention relates to scalable encoding and decoding of a video signal, and more particularly, to a method and apparatus for encoding a video signal, which further uses the base layer of the video signal to code an enhanced layer of the video signal, and to decode the encoded video data. A method and apparatus are disclosed.

A scalable video codec (SVC) allows a portion of an encoded picture sequence (particularly a sequence of frames intermittently selected from the entire sequence of frames) to be decoded and used to represent a low quality video, resulting in a high quality picture. This is a method of encoding into a sequence. Motion Compensated Temporal Filtering (MCTF) is an encoding scheme proposed to be used for scalable video codecs.

Although the low quality video can be represented by receiving and processing a part of the scalable picture sequence as described above, there is still a problem that the image quality is greatly reduced when the bit rate is low. As a solution to this problem, hierarchically provide a secondary picture sequence for low bitrate, e.g., a picture sequence with a small screen size and / or a low frame rate, so that each decoder can create a sequence suitable for its capabilities and characteristics. To decode it. For example, the secondary picture sequence of the CIF and the secondary picture sequence of the QCIF (Quarter CIF) as well as the primary picture sequence of 4CIF (common intermediate format) are encoded and transmitted to the decoder. Each sequence is called a layer, the higher of the two given layers is called the enhanced layer and the lower layer is called the base layer.

This picture sequence has redundancy because the same video signal source is encoded into the sequences. In order to increase the coding efficiency of each sequence, it is possible to reduce the amount of high sequence coding information by performing inter-sequence picture prediction of high sequence video frames from low sequence video frames that match the high sequence video frames in time. Is necessary.

However, video frames of sequences of different layers may have different aspect ratios. For example, a high sequence (ie, enhanced layer) video frame may have a large aspect ratio of 16: 9, while a low layer (ie, base layer) video frame may have a small aspect ratio of 4: 3. In this case, when performing the prediction of the enhanced layer picture, it is necessary to determine which part of the base layer picture is used for the enhanced layer picture or which part of the enhanced layer picture is used.

Summary of the Invention

The present invention relates to the decoding of video signals.

In one embodiment of a method of decoding a video signal, at least a portion of the current image in the current layer is predicted based on at least a portion of the reference image, offset information and dimension information. The offset information may indicate a position offset between at least one boundary pixel of the reference image and at least one boundary pixel of the current image, and the dimension information may indicate at least one dimension of the current image.

In one embodiment, the reference image is based on the base image in the base layer. For example, the reference image can be at least an upsampled portion of the base image.

In one embodiment, the offset information includes left offset information indicating a position offset between at least one left pixel of the reference image and at least one left pixel of the current image.

In another embodiment, the offset information includes top side offset information indicating a position offset between at least one top side pixel of the reference image and at least one top side pixel of the current image.

In one embodiment, the dimension information includes width information indicating the width of the current image.

In another embodiment, the dimension information includes height information indicating the height of the current image.

In one embodiment, the offset information may be obtained from the header for at least a portion (eg, slice, frame, etc.) of the picture in the current layer. It is also possible to determine that there is offset information based on the indicator in the header.

Another embodiment includes an apparatus for decoding a video signal.

1 is a block diagram illustrating a video signal encoding apparatus to which a scalable video signal coding method according to the present invention is applied.

FIG. 2 is a block diagram illustrating a portion of the MCTF encoder shown in FIG. 1 performing image estimation / prediction and update operations.

3A and 3B show a relationship between an enhanced layer frame and a base layer frame that can be used as a reference frame for converting an enhanced layer frame into an H frame having a predictive image.

4 is a diagram illustrating a method in which a portion of a base layer picture is selected and enlarged and used for a prediction operation of an enhanced layer picture according to an embodiment of the present invention.

5A and 5B illustrate an embodiment of the structure of information about the positional relationship between a base layer picture and an enhanced layer picture, which is transmitted to a decoder in accordance with the present invention.

FIG. 6 illustrates a method in which an area including a base layer picture is enlarged and used for a prediction operation of an enhanced layer picture according to another embodiment of the present invention; FIG.

7 illustrates a method in which a base layer picture is enlarged to an area larger than an enhanced layer picture and used for a prediction operation of an enhanced layer picture according to another embodiment of the present invention.

8 is a block diagram illustrating an apparatus for decoding a data stream encoded by the apparatus of FIG.

FIG. 9 is a block diagram illustrating a portion of the MCTF decoder shown in FIG. 8 performing inverse prediction and update operations. FIG.

Hereinafter, with reference to the accompanying drawings will be described an embodiment of the present invention;

1 is a block diagram illustrating a video signal encoding apparatus to which a scalable video signal coding method according to the present invention is applied. Although the apparatus of FIG. 1 is implemented to code two layers of input video signal, the principles of the present invention described below can be applied when a video signal is coded in three or more layers. The invention is also not limited to the MCTF scheme described below by way of example, but may be applied to any scalable video coding scheme.

The video signal encoding apparatus shown in FIG. 1 includes an MCTF encoder 100, a texture coding unit 110, a motion coding unit 120, and a base layer encoder to which the present invention is applied. (BL Encoder) 150 and Muxer (Multiplexer). The MCTF encoder 100 is an enhanced layer encoder that encodes an input video signal per macroblock and generates appropriate management information according to the MCTF scheme. The texture coding unit 11 converts the information of the encoded macroblock into a compressed bitstream. The motion coding unit 120 converts the motion vector of the image block obtained by the MCTF encoder 100 into a compressed bitstream according to a specific scheme. The base layer encoder 150 encodes the input video signal according to a specific method, for example, the MPEG-1, 2 or 4 standard or the H.261, H.263 or H.264 standard, and the picture sequence of the small picture, For example, create a picture sequence reduced to 25% of its original size. The multiplexer 130 encapsulates the output data of the texture coding unit 110, the picture sequence of the small screen output from the base layer encoder 150, and the output vector data of the motion coding unit 120 into a desired format. The multiplexer 130 multiplexes and encapsulates the encapsulated data in a desired transmission format. The base layer encoder 150 not only encodes the input video signal into a picture sequence having a smaller picture size than the enhanced layer picture, but also encodes the input video signal into a picture sequence having the same picture size as the enhanced layer picture at a lower frame rate than the enhanced layer picture. By encoding, a low bitrate data stream can be provided. In the embodiment of the present invention described below, the base layer is encoded into a picture sequence of a small screen, the picture sequence of the small screen is called a base layer sequence, and the frame sequence output from the MCTF encoder 100 is called an enhanced layer sequence. .

The MCTF encoder 100 performs motion estimation and prediction on each target macroblock in the video frame. The MCTF encoder 100 performs an update operation on each target macroblock by adding an image difference between the target macroblock and the corresponding macroblock in the neighboring frame to the corresponding macroblock in the neighboring frame. 2 shows some elements of the MCTF encoder 100 performing these operations.

Elements of the MCTF encoder 100 shown in FIG. 2 include an estimator / estimate 102, an update 103, and a decoder 105. The decoder 105 decodes the encoded stream received from the base layer encoder 150 and enlarges the decoded small picture frame to the frame size of the enhanced layer using the internal scaler 105a. The estimator / predictor 102 retrieves the reference block of each macroblock in the current frame to be coded with the residual data, in the adjacent frame before or after the current frame, and in the frame enlarged by the scaler 105a. The estimator / predictor 102 obtains the image difference (ie, inter-pixel difference) of each macroblock in the current frame from the corresponding block in the temporally matched frame enlarged by the reference block or scaler 105a and macros the image difference. Coding into blocks The estimator / predictor 102 also obtains a motion vector extending from the macroblock to the reference block. The updater 103 multiplies the image difference of the macroblock by an appropriate constant (e.g., 1/2 or 1/4) and adds the resulting value to the reference block to thereby reference within the frame before or after the current frame. Performs an update operation of the macroblock in the current frame in which the block is found. The operation performed by the updater 103 is called "U" operation and the frame generated by the "U" operation is called "L" frame.

The estimator / predictor 102 and updater 103 of FIG. 2 may perform operations on a plurality of slices generated by dividing a single frame simultaneously or in parallel, instead of performing operations on video frames. Frames (or slices) with image differences generated by estimator / predictor 102 are referred to as "H" frames (or slices). An "H" frame (or slice) includes data with high frequency components of a video signal. In the following description of embodiments, the term "picture" is used to refer to a slice or frame, if the use of that term is technically possible.

The estimator / predictor 102 splits each of the input video frames (or L frames obtained at the previous level) into macroblocks of the desired size. For each segmented macroblock, the estimator / predictor 102 determines that the image is very much like each segmented macroblock in the base layer frame magnified by the previous / next neighbor frame of the enhanced layer and / or scaler 105a. Search for similar blocks. That is, the estimator / predictor 102 retrieves macroblocks that are correlated in time with each divided macroblock. A block having an image very similar to the target image block has the smallest image difference from the target image block. The image difference between two image blocks is defined as, for example, the sum or average of the difference between pixels of two image blocks. Among the blocks having a smaller or critical image difference than the target macroblock in the current frame, the block having the smallest image difference with the target macroblock is called a reference block. A picture including a reference block is called a reference picture. For each macroblock of the current frame, two reference blocks (or two reference pictures) exist in the previous frame of the current frame (including the base layer frame) or the next frame of the current frame (including the base layer frame), There may be one reference block in the previous frame and one reference block in the next frame.

When the reference block is found, the estimator / predictor 102 calculates and outputs a motion vector from the current block to the reference block. The estimator / predictor 102 may also determine the pixel error value of the current block (ie, the pixel difference value) from the pixel value of the reference blow present in the previous frame or the next frame or the average pixel value of two reference frames present in the previous and next frame. ) And output. Image or pixel difference values are also referred to as residual data.

If a macroblock having a desired threshold image difference or smaller than the current macroblock cannot be found in two neighboring frames (including the base layer frame) through a motion estimation operation, the estimator / predictor 102 is in the same time domain as the current frame. It is determined whether a frame (hereinafter, referred to as a temporal coincidence frame) or a frame in a time domain (hereinafter referred to as a temporal proximity frame) close to the current frame exists in the base layer sequence. If such a frame is present in the base layer sequence, the estimator / predictor 102 derives the image difference (i.e., residual data) of the current macroblock from the corresponding macroblock in the temporal match or proximity frame based on the pixel values of the two macroblocks. The motion vector of the current macroblock is not obtained for the corresponding macroblock. The proximity time zone for the current frame corresponds to a time interval that includes a frame that can be considered to have the same image as the current frame. The information of this time interval is included in the encoded stream.

The above-described operation of the estimator / predictor 102 is referred to as "P" operation. If estimator / predictor 102 performs a " P " operation by retrieving a reference block of each macroblock in the current frame and coding each macroblock with residual data, estimator / predictor 102 generates an H frame. As the reference picture, as shown in FIG. 3A, in addition to the neighboring L frames of the enhanced layer before and after the current frame, an enlarged picture of the base layer received from the scaler 105a may be selectively used.

In one embodiment of the invention, five frames are used to generate each H frame. 3B shows five frames that can be used to generate an H frame. As shown, the current L frame 400L has an L frame 401 before the current L frame 400L and an L frame 402 after the current L frame 400L. In addition, the current L frame 400L has a base layer frame 405 within the same time domain. One or two of the L frames 401 and 402 of the same MCTF level as the current L frame 400L, before the frame 405 and the frame 405 of the base layer in the same time domain as the L frame 400L and Subsequent base layer frames 403 and 404 are used as reference pictures to generate an H frame 400H from the current L frame 400L. As mentioned above, there are various reference block selection modes. In order to inform the decoder which mode is selected, the MCTF encoder 100 inserts "reference block selection mode" information into a field at a specific position of the header area of the corresponding macroblock and then textures the "reference block selection mode" information. Transmission to the coding unit 110.

If the picture of the base layer is used as a reference picture for picture prediction of the enhanced layer in the reference picture selection method as shown in FIG. 3B, all or part of the base layer picture may be used for prediction of the enhanced layer picture. For example, as shown in FIG. 4, the base layer picture has an aspect ratio of 4: 3, the actual image portion 502 of the base layer picture has a 16: 9 aspect ratio, and the enhanced layer picture 500 has 16 With an aspect ratio of: 9, the upper and lower horizontal portions 501a and 501b of the base layer picture contain invalid data. In this case, only the image unit 502 of the base layer picture is used for prediction of the enhanced layer picture 500. To achieve this, the scaler 105a selects (or cuts) the image portion 502 of the base layer picture (S41), and up-samples the selected image portion 502 to enhance the enhanced layer picture ( In step S42, the enlarged image unit is provided to the estimator / predictor 102.

The MCTF encoder 100 merges the position information of the selected portion of the base layer picture into the header of the current picture coded with the residual data. The MCTF encoder 100 also sets a flag "flag_base_layer_cropping" indicating that a portion of the base layer picture is selected and used to insert it into the picture head at the appropriate location so that the flag is passed to the decoder. Position information is not transmitted when the flag "flag_base_layer_cropping" is reset.

5A and 5B show an embodiment of the structure of the information about the selected portion 512 of the base layer picture. In the embodiment of FIG. 5A, selected portions 512 of the base layer picture are designated by offsets (left_offset, right_offset, top_offset and bottom_offset) from the left, right, top and bottom boundaries of the base layer picture. The left offset represents the position offset between the left pixel (or, for example, at least one pixel) in the base layer image and the left pixel in the selected portion 512. The top side offset represents the position offset between the top side pixel (or, for example, at least one pixel) in the base layer image and the top side pixel in the selected portion 512. The right offset represents the position offset between the right pixel (or, for example, at least one pixel) in the base layer image and the right pixel in the selected portion 512. The bottom side offset represents a position offset between bottom side pixels (or, for example, at least one pixel) in the base layer image and bottom side pixels in the selected portion 512. In the embodiment of FIG. 5B, the selected portion 512 of the base layer picture is determined by offsets (left_offset and top_offset) from the left and top boundaries of the base layer picture and the width and height (crop_width and crop_height) of the selected portion 512. Is specified. In addition, other various designation methods are possible.

The offset in the information of the selected portion shown in Figs. 5A and 5B may have a negative value. For example, as shown in FIG. 6, the base layer picture has an aspect ratio of 4: 3, the enhanced layer picture 600 has an aspect ratio of 16: 9, and the actual image portion of the picture has a 4: 3 aspect ratio. If left, the left and right offset values left_offset and right_offset have negative values -d _L and -d _R. Portions 601a and 601b extending from the base layer picture are designated by negative values -d _L and -d _R. The extended portions 601a and 601b are filled with offscreen data, and the picture 610 including the extended portions 601a and 601b is upsampled to have the same size as the enhanced layer picture 600. Thus, data of the region 611 in the enlarged bay layer picture corresponding to the actual image portion of the enhanced layer picture 600 may be used for prediction of the actual image portion of the enhanced layer picture 600.

Since the offset field of the information shown in FIGS. 5A and 5B may have a negative value, instead of using the information of FIGS. 5A and 5B specifying a selected area within the base layer picture, the enhanced layer picture associated with the enlarged base layer picture and The same advantages as those described in the example of FIG. 4 are obtained by using the information of FIGS. 5A and 5B as positional information of overlapping regions.

In particular, referring to FIG. 7, when the base layer picture 702 is upsampled and the actual image area 701 of the base layer picture 702 is enlarged to the size of the enhanced layer picture 700, it is enlarged (eg, The, upsampled) picture corresponds to an area larger than the enhanced layer picture 700. In this example, the top and bottom offsets top_offset and bottom_offset are included in the positional information of the region overlapping with the enhanced layer picture 700. These offsets correspond to the enlarged base layer picture and negative values (-d _T and d _B ) are assigned such that only the actual image area of the enlarged base layer picture is used for prediction of the enhanced layer picture 700. In the example of FIG. 7, the left and right offsets of the position information of the area corresponding to the enlarged base layer picture are zero. However, it will be appreciated that the left and right offsets may not be zero and correspond to the enlarged base layer picture. Some of the images in the enlarged base layer picture may not be used to determine the enhanced layer picture. Similarly, if the offset information corresponds to the base layer picture, a portion of the image of the base layer picture may not be used to determine the enhanced layer picture, as opposed to the upsample base layer picture.

Also, in this embodiment, the left offset represents the position offset between the left pixel (or, for example, at least one pixel) in the upsampled base layer image and the left pixel in the enhanced layer image. The top side offset represents the position offset between the top side pixel (or at least one pixel, for example) in the upsampled base layer and the top side pixel in the enhanced layer image. The right offset represents the position offset between the right pixel (or at least one pixel, for example) in the upsampled base layer image and the right pixel in the enhanced layer image. The bottom side offset represents a position offset between bottom side pixels (or, for example, at least one pixel) in the upsampled base layer image and bottom side pixels in the enhanced layer image.

As described above, the information in FIGS. 5A and 5B may be used as information for selecting a part of a base layer picture used for prediction of an enhanced layer picture or an enhanced layer picture associated with a base layer picture used for prediction of an enhanced layer picture. It can be used as location information of overlapping regions.

Information of the size and aspect ratio of the base layer picture, mode information of the actual image of the base layer picture, and the like can be determined, for example, by decoding from the sequence header of the encoded base layer stream. That is, the information may be recorded in the sequence header of the encoded base layer stream. Thus, the position of the region overlapping the enhanced layer picture corresponding to the selected region or base layer picture in the above-described base layer picture is determined based on the position or offset information, and all or part of the base layer picture is adapted to adapt this determination. Used.

1 and 2, the MCTF encoder 100 performs the above-described " P " and " U " operations on a picture sequence of a predetermined length, for example, a picture group (GOP), so that the sequence of the H frame and the L frame are performed. Create a sequence of. The estimator / predictor and updater of the next stage (not shown) connected in series then generates a sequence of H frames and a sequence of L frames by repeating the "P" and "U" operations on the generated L frame sequence. . The "P" and "U" operations are performed a predetermined number of times (eg, until one L frame is generated for a GOP) to produce the final enhanced layer sequence.

The data stream encoded in the above-described manner is transmitted wirelessly or by wire to the decoding apparatus or via a recording medium. The decoding apparatus reconstructs the original video signal in the enhanced and / or base layer according to the method described below.

8 is a block diagram illustrating an apparatus for decoding a data stream encoded by the apparatus of FIG. 1. The decoding apparatus of FIG. 8 includes a demux (demultiplexer) 200, a texture decoding unit 210, a motion decoding unit 220, an MCTF decoder 230, and a base layer decoder 240. The demultiplexer 200 separates the received data stream into a compressed motion vector stream, a compressed macroblock information stream, and a base layer stream. The texture decoding unit 210 reconstructs the compressed macroblock information stream to an uncompressed original state. The motion decoding unit 220 reconstructs the compressed motion vector stream to an uncompressed original state. The MCTF decoder 230 is an enhanced layer decoder for converting an uncompressed macroblock information stream and an uncompressed motion vector stream into an original video signal according to the MCTF scheme. Base layer decoder 240 decodes the base layer stream according to a specified scheme, for example, according to the MPEG-4 or H.264 standard.

The MCTF decoder 230 includes an inverse filter having the structure shown in FIG. 9 for reconstructing an input stream into an original frame sequence as an internal element.

9 shows some elements of an inverse filter that reconstruct a sequence of H and L frames of MCTF level N into a sequence of L frames of level N-1. The elements of the inverse filter of FIG. 9 include an inverse updater 231, an inverse predictor 232, a motion vector decoder 235, an arranger 234, and a scaler 230a. The inverse updater 231 subtracts the pixel difference value of the input H frame from the corresponding pixel value of the input L frame. The inverse predictor 232 reconstructs the input H frame into a frame having the original image with reference to the L frame from which the image difference of the H frame is subtracted and / or the enlarged picture output from the scaler 240a. The motion vector decoder 235 decodes the input motion vector stream into motion vector information of each block and supplies the motion vector information to the inverse predictor (eg, the inverse predictor 232) of each stage. The arranger 234 interleaves the frames completed by the inverse predictor 232 between L frames output from the inverse updater 231 to generate a normal video frame sequence. The scaler 230a enlarges the small screen picture of the base layer to the size of the enhanced layer picture according to the information as shown in FIGS. 5A and 5B, for example.

The L frame output from the arranger 234 constitutes an L frame sequence 601 of level N-1. The inverse updater and predictor next to level N-1 reconstruct L frame sequence 601 and input H frame sequence 602 at level N-1 into L frame sequences. This decoding process is performed the same number of times as the number of MCTF levels employed in the encoding process to reconstruct the original video frame sequence. With reference to " reference_selection_code " information contained in the header of each macroblock of the input H frame, the inverse predictor 232 designates the macro by specifying an enlarged frame of the base layer and / or an L frame of the enhanced layer used as the reference frame. Code the block with residual data. The inverse predictor 232 determines a reference block within a specified frame based on the motion vector provided from the motion vector decoder 235 and determines the pixel value of the reference block (or the average pixel value of two macroblocks used as the reference block of the macroblock). ) Is added to the pixel difference value of the macroblock of the H frame to reconstruct the original image of the macroblock of the H frame.

If the base layer picture is used as a reference frame of the current H frame, the scaler 230a selects and enlarges an area within the base layer picture (in the example of FIG. 4) or contains a picture contained in a header analyzed by the MCTF decoder 230. Based on the positional relationship information as shown in 5a and 5b, the area larger than the base layer picture (in the example of FIG. 6) is enlarged so that the enlarged area of the base layer picture is used to display the remaining data in the current H frame as described above. It is used to reconstruct the containing macroblock into the original image block. The position relationship information is extracted from the header and is referred to when the information indicating whether the position relationship information is included (in particular, the flag "flag_base_layer_cropping" in the example of FIGS. 5A and 5B) indicates that the position relationship information is included.

When the information of FIGS. 5A and 5B is used as the information indicating the position of the region overlapping with the enhanced layer picture used for the prediction of the enhanced layer picture, the inverse predictor 232 is enlarged according to the value (positive or negative) of the offset information. The enlarged one of the base layer pictures received from the scaler 230a for prediction of the enhanced layer picture is used by associating the entirety of the base layer picture with an area larger than the current H frame or all or a portion of the current H frame. In the case of FIG. 7 in which the enlarged base layer picture is larger than the current H frame, the predictor 232 enlarges the base layer corresponding to the H frame to reconstruct the macroblock in the current H block into the original image. Only the area of the picture is used. In this example, the offset information includes a negative value.

For one H frame, MCTF decoding is performed in specific units, for example in parallel slice units, so that the macroblocks within the frame have their own original image reconstructed, and the reconstructed macroblocks are combined to form a complete video frame. Configure

The decoding method described above reconstructs the MCTF encoded data stream into a complete video frame sequence. The decoding apparatus decodes and outputs a base layer sequence using a base layer that depends on its processing and presentation capability, or decodes and outputs an enhanced layer sequence.

The above-described decoding apparatus may be incorporated into a mobile communication terminal, a media player, or the like.

As can be seen from the above description, the method and apparatus for encoding / decoding a video signal according to the invention, in addition to the picture of the enhanced layer, when encoding the video signal in a scalable manner such that the total amount of coded data is reduced. Coding efficiency is increased by using the picture of the base layer provided to the low performance decoder. In addition, even when a portion of the base layer picture that can be used for the prediction operation of the enhanced layer picture is specified and a picture enlarged from the base layer picture cannot be directly used for the prediction operation of the enhanced layer picture, the prediction operation may be normally performed without degrading performance. have.

While the invention has been described with reference to exemplary embodiments, it will be apparent to those skilled in the art that various modifications, changes, substitutions, and additions can be made without departing from the scope of the invention. Accordingly, the present invention includes improvements, modifications, substitutions and additions to the present invention.

Claims (26)

Obtaining location information of a current block in a current image in the enhanced layer; Obtaining offset information between at least one boundary pixel of the reference image in the base layer and at least one boundary pixel of the current image; Acquiring position information of a reference block in the reference image in the base layer based on the position information and the offset information of the current block; Obtaining reference residual data based on the positional information of the reference block; Obtaining residual data of the current block based on the reference residual data; And Decoding the current block based on the remaining data of the current block Video signal decoding method comprising a. The method of claim 1, The enhanced layer has a screen ratio or spatial resolution different from that of the base layer, and the base layer corresponds to the same video signal as the enhanced layer. delete The method of claim 1, And the reference image represents an upsampled portion of a base layer image. The method of claim 1, And the reference residual data represents an upsampled portion of a base layer image. The method of claim 5, And the upsampled portion of the base layer image is residual data coded. The method of claim 1, The offset information is, Left offset information indicating a position offset between at least one left side pixel of the reference image and at least one left side pixel of the current image; Upper offset information indicating a position offset between at least one upper side pixel of the reference image and at least one upper side pixel of the current image; Right offset information indicating a position offset between at least one right side pixel of the reference image and at least one right side pixel of the current image; Lower offset information indicating a position offset between at least one lower side pixel of the reference image and at least one lower side pixel of the current image Video signal decoding method comprising a. The method of claim 1, Determining whether the offset information is obtained from a header of at least one portion of the current image, And said offset information is obtained based on said determining step. The method of claim 1, Upsampling at least a portion of a base layer image to obtain an upsampled image as the reference image. The method of claim 9, Wherein said portion of said base layer image represents a residual data coded portion of a base layer image. The method of claim 1, And the position information of the reference block is obtained based on size information indicating at least one size of the current image. The method of claim 11, The size information includes width information indicating a width of the current image and height information indicating a height of the current image. delete The method of claim 1, And the offset information is obtained from a slice header of the enhanced layer. A demux for obtaining position information of a current block in a current image in an enhanced layer and offset information between at least one boundary pixel of a reference image in a base layer and at least one boundary pixel of the current image; A decoder that obtains position information of a reference block in the reference image in the base layer based on the position information and the offset information of the current block, and obtains reference residual data based on the position information of the reference block Video signal decoding apparatus comprising a. delete delete delete delete delete delete delete delete delete delete delete

Images