KR101549017B1 - Image codec system, image encoding method, and image decoding method for supporting spatial random access - Google Patents

Abstract

The present invention proposes an image codec system, an image encoding method, and an image decoding method that support spatial random access. The image codec system according to an embodiment of the present invention generates two or more bit streams for each piece image by sequentially encoding each of the plurality of fragment images in two or more scanning directions having different starting positions, A first encoding server for extracting and transmitting one bit stream among the two or more bit streams of each piece image having an area overlapping the target image based on the position information of the image and a bit stream transmitted from the first encoding server And decodes the image in the scanning direction to reconstruct a target image. According to the present invention, after each piece image is encoded in various scanning directions, considering the relative position of the target image with respect to each piece image, encoding is performed according to the selected scanning direction so as to minimize the amount of transmission data And extract and transmit only bitstreams for all or part of each piece image, thereby improving transmission and decoding efficiency.

Description

[0001] The present invention relates to an image codec system, an image encoding method, and a decoding method for supporting spatial random access,

The present invention relates to an image codec system, and more particularly, to an image codec system, an image encoding method, and an image decoding method that support spatial random access.

Recently, digitalization of broadcasting and fusion of broadcasting and communication are actively proceeding. Accordingly, the broadcasting environment has been changed into a complex form in which bidirectional communication is possible in the conventional unidirectional reception, and interworking of heterogeneous networks and accommodating various terminals. A typical example that reflects this trend is ubiquitous television (TV). Ubiquitous TV refers to a TV that can be viewed whenever and wherever a digital device is connected to the Internet. To watch ubiquitous TV, you connect a tuner for ubiquitous TV to the antenna of cable or terrestrial TV and connect it to high-speed internet line. By using such a ubiquitous TV, it is possible to remotely control a ubiquitous TV tuner using a portable terminal such as a notebook computer, a personal digital assistant (PDA) or a WiBro terminal, so that a variety of media such as movies, can see. The service using the ubiquitous TV is rapidly emerging as a next-generation service because the user can receive desired information anytime and anywhere by using a terminal of a desired form.

  However, the ubiquitous environment that can share information in various terminals is creating a new problem of resolution inconsistency between terminals. For example, the latest high definition television (HDTV) supports resolutions up to 1920 × 1080, and the movie-specific digital format newly planned by the US-based movie industry is based on 4K × 2K digital cinema . On the other hand, PDAs, portable multimedia players (PMPs) and WiBro handsets are portable, so they only support resolutions up to CIF (Common Intermediate Format) or QCIF (Quater Common Intermediate Format).

One method for solving the above-mentioned problem of resolution mismatch is to down-sample a high-resolution image received by a TV or the like, and then transmit a low-resolution image to the mobile terminal. However, since the downsampled low-resolution image provides a low-quality screen to the user as compared with the case of directly watching the TV, the user can not satisfy the desire of the user who wants to view the high-quality screen through the portable terminal. A user using a device such as a music video, a sports game such as a soccer, a home shopping, a navigation, or the like may wish to view a specific part of an image of interest in detail. In the above-described downsampling method, It is difficult to satisfy.

One of the proposed methods to solve this problem is Spatial Random Access. The spatial random access is intended to enable transmission, restoration and retrieval of only a part of an image at an arbitrary position in the image. If this is used, not only the data related to the whole image but also a specific part considering the user's intention and the size of the display window It is possible to perform partial decoding by transmitting only the data related to the data. Therefore, by using an image codec system supporting spatial random access, it becomes possible to view a high-quality image such as an HDTV with a portable terminal.

Due to diversification and popularization of portable terminals, there is an increasing need for an image codec system supporting this random access. In one of the subgroups of MPEG, currently, free-view-point TV (F-TV) and three-dimensional TV are used in relation to multi-view video coding (MVC) Target Application). As a result of the 76th MPEG Swiss Meeting held in April 2006, the Free View Scalability, which makes only a part of the image accessible, It came out as a proposal. The pre-view scalability is a technique capable of decoding only a portion of each camera in multi-view video coding.

In addition, even in the case of a panoramic image created to provide a wider field of view to the user, it is considered that such a random access should be applied to the image. Since the panoramic image is a reconstructed image of a single image taken at various viewpoints, the size of the panoramic image is increased to several tens times as large as that of a general image. If a part of the panorama image can be transmitted and decoded by using an image codec system supporting spatial random access instead of transmitting and decoding the entire panorama image, it is possible to not only satisfy a user's desire to view a high-quality image, Which is quite effective in the present invention.

However, it is difficult to realize this spatial random access efficiently with the video codec system and image coding / decoding method that have been developed or proposed so far. Since there is no need for broadcast or communication services in which various terminals coexist, there have been no extensive researches on specific methods for realizing this random access. In recent years, however, research has been progressing slowly. Representative examples include "Efficient Representation and Interactive Streaming" proposed by Carsten Grnheit et al. (Proc. ICIP 2002, IEEE International Conference on Image Processing, Rochester, -25, 2002).

In this paper, we propose a method of dividing a whole image into several pieces of image and encoding them in units of pieces. Each piece of the encoded piece of image can be transmitted and decoded in units of pieces of image. Therefore, considering the size of the display of the user and the area (target image) desired by the user in the entire image, only the fragment image necessary for displaying the target image can be transmitted and decoded.

FIGS. 1A to 1C are diagrams showing the relationship between the size of a piece image and the area to be encoded and transmitted in case of using the method disclosed in the above paper. Referring to FIGS. 1A to 1C, it can be seen that as the size of the target image is constant, the smaller the unit size of the sculptured image is, the smaller the transmission area to be encoded and transmitted is. This is because the method presented in the above paper encodes and transmits all of the sculptured images even when the target image overlaps part of the sculptured image. As shown in FIG. 1C, if the size of the sculptured image is made very small, the size of the target image and the size of the transmission area become substantially similar.

However, general image coding techniques use decoded data of neighboring blocks located around the current block to be encoded in order to increase the compression ratio. When blocks are sequentially encoded in accordance with a raster scan sequence that is commonly used, reconstructed reconstructed data of blocks that are located at the left and above the current block to be encoded are generally used. Therefore, as the unit size of the sculptured image becomes smaller, the reconstruction data of the neighboring blocks can not be used by that much, and the encoding efficiency is lowered accordingly. As a result, it is not preferable to reduce the unit size of the sculptured image to a certain size or smaller (for example, the size as shown in FIG. 1C) in consideration of the coding efficiency.

Therefore, in order to increase the coding efficiency, it is preferable to set the size of the piece image to a predetermined size or more as in the case of FIG. 1A or FIG. 1B. However, when encoding is performed according to the method disclosed in the above paper, another problem arises when the size of the fragment image increases. More specifically, even when a target image overlaps with a part of a specific piece of image for reasons such as spanning several pieces of image, there arises a problem of encoding, transmitting, and decoding the whole fragmented image. This problem becomes more serious as the size of the sculptured image is larger and / or the target image overlaps with a smaller portion of the sculptured image.

The reason for such a problem is that the conventional general image encoding techniques including the method disclosed in the above paper sequentially encode blocks from the upper left corner of each piece image according to the raster scan direction. The raster scan direction is shown in FIG. 2, in which one row of blocks is encoded from the leftmost block to the right, and then the rightmost block is encoded from the leftmost row to the right. The macroblocks encoded in the raster scan direction must also be decoded in the raster scan direction. As a result, if the target image is located on the lower right side of the sculptured image, encoding, transmission and decoding for unnecessary portions not included in the target image are increased. In the worst case, even if the target image covers only one macroblock at the bottom right of the sculptured image, there is a problem that the sculptured image as a whole must be encoded, transmitted, and decoded.

As described above, in the conventional image coding method that divides the whole image into a sculptured image and uniformly scans all the sculptured images according to the raster scan direction, the size of the sculptured image must be reduced to a small extent to improve the efficiency of the random access. However, reducing the size of the sculptured image has a problem in that the efficiency of encoding, transmission and decoding is greatly reduced. In addition, if the size of the sculptured image is increased and all the sculptured images are encoded according to one scanning direction, there arises a problem of transmission and decoding of unnecessary areas in the case of a specific sculptured image.

In order to solve these problems, the present invention solves the above-described problems of the prior art related to the spatial random access method, thereby improving coding and transmission efficiency, efficiently using a limited transmission channel, We propose an image codec system, an image coding method and a video decoding method that support a spatial random access method for viewing high resolution images.

According to an aspect of the present invention, there is provided a system for supporting a spatial random access to a target image, which is a part of a whole image partitioned into a plurality of sculptured images, By sequentially encoding the blocks in accordance with two or more scan directions having different start positions, thereby generating two or more bit streams for each of the piece images, and based on the position information of the target images, A first encoding server for extracting one bitstream from among the two or more bitstreams of each of the piece images having an overlapping region, and a second encoding server for extracting the bitstream transmitted from the first encoding server, And a reconstructing unit that reconstructs the object image .

According to an aspect of the embodiment, the first encoding server includes an encoder unit for generating the two or more bit streams for each piece image, a memory unit for storing a bit stream generated in the image encoding apparatus, And a bitstream extraction unit for extracting one bitstream from the two or more bitstreams of each piece image stored in the memory unit. According to another aspect of the present invention, the image codec system generates a bit stream by encoding each piece image having an area overlapping the target image based on position information of the target image according to a raster scan direction, And a second encoding server for transmitting the bitstream, wherein the decoding apparatus decodes the bitstream transmitted from the second encoding server according to a raster scan direction to reconstruct the target image. In this case, the location information of the target image may be transmitted from the decoding apparatus to the first encoding server or the second encoding server.

According to another aspect of the present invention, there is provided a system for supporting spatial random access to a target image, which is a part of an entire image partitioned into a plurality of rectangular piece images, Generates four bit streams for each of the fragmented images by coding each of the blocks adjacent to four vertexes of the rectangle as a start position and encoding the blocks adjacent to the start position in accordance with a scan direction for sequentially encoding the blocks, A first encoding server for extracting one bit stream among the four bit streams of each piece image having an area overlapping with the target image based on position information of the image, Decodes the stream in accordance with the scanning direction, And a video decoding apparatus for reconstructing the image.

According to an aspect of the embodiment, the first encoding server includes an encoder unit for generating the four bit streams for each piece image, a memory unit for storing a bit stream generated in the image encoding apparatus, And a bitstream extractor for selecting and extracting one bitstream among the four bitstreams of each piece image stored in the memory unit.

According to another aspect of the present invention, the image codec system generates a bit stream by encoding each piece image having an area overlapping the target image based on position information of the target image according to a raster scan direction, And a second encoding server for transmitting the bitstream, wherein the decoding apparatus decodes the bitstream transmitted from the second encoding server according to a raster scan direction to reconstruct the target image.

According to still another aspect of the present invention, the four bitstreams are arranged such that a block adjacent to the upper left vertex of the rectangle is used as a start position, (1) blocks of a row are sequentially encoded from left to right, A method of coding blocks sequentially from the left to the right in a lower row and a method of sequentially coding blocks in a column from top to bottom and sequentially coding blocks from the top to bottom in the right column A first bit stream coded by the first bitstream, a block adjacent to the upper right corner of the rectangle as a start position, (1) sequentially encoding a block of one row from right to left, and A method of sequential coding and a method of A second bitstream coded by any one of the following methods: a block adjacent to the vertex at the lower left of the rectangle is set as a start position, A method of successively encoding a block of a row to the right in the upper row and sequentially encoding the blocks in a row from the left to the right of the upper row, and (2) a method of sequentially encoding one row of the block from the bottom up, And a block adjacent to the lower right corner of the rectangle as a start position, and (1) a block of one row from the right to the left in a sequential manner , And again the upper row A method of sequentially encoding blocks sequentially from the bottom to the top, and a method of sequentially encoding blocks sequentially from the bottom to the top and sequentially encoding the blocks from the bottom to the top of the left column. Bit stream. In this case, if the target image includes the upper right corner of the fragment image among the four bit streams of each fragment image, the first encoding server converts the first bit stream, the target image, The second bitstream, the third bitstream if the target image includes the lower left vertex of the sculptured image, or the lower right vertex of the sculpted image. The fourth bitstream can be extracted.

According to another aspect of the present invention, there is provided an image codec system for supporting a spatial random access to a target image, which is a part of an entire image partitioned into at least two rectangle slice images by at least one vertical boundary line Wherein each of the two or more piece images is encoded in accordance with a scanning direction in which blocks adjacent to the top two vertexes of the rectangle are sequentially set as starting positions and blocks adjacent to the starting position are sequentially encoded, A first encoding server for generating one bit stream from among the two bit streams of each piece image having an area overlapping the target image based on the position information of the target image, And a second encoding server And an image decoding apparatus for decoding the trim in accordance with the scan direction to reconstruct the target image.

According to an aspect of the present invention, there is provided a method of supporting spatial random access to a target image, which is a part of a whole image partitioned into a plurality of fragmented images, Generating two or more bit streams for each piece image by sequentially encoding the blocks in accordance with two or more scanning directions having different starting positions, and generating a bit stream corresponding to a region overlapping the target image based on the position information of the target image And extracting one bitstream from the two or more bitstreams of each of the piece images.

According to another aspect of the present invention, there is provided a method of supporting spatial random access to a target image, which is a part of a whole image partitioned by a plurality of rectangular piece images, Generating four bit streams for each piece image by coding each of blocks adjacent to four vertexes of the rectangle as a start position and encoding the blocks adjacent to the start position in accordance with a scan direction for sequentially encoding the blocks; Extracting one bit stream from among the four bit streams of each piece image having an area overlapping the target image based on position information of the target image.

According to another aspect of the present invention, there is provided a method of encoding an image, the method comprising: providing spatial random access to a target image, which is a part of a whole image partitioned into at least two rectangular slice images by one or more vertical boundaries; A method for encoding an image, the method comprising: encoding each of the two or more piece images in accordance with a scanning direction in which blocks adjacent to the top two vertexes of the rectangle are set as starting positions and blocks adjacent to the starting position are sequentially encoded; Extracting one bit stream from the two bit streams of each piece image having an area overlapping the target image based on the position information of the target image, .

According to an aspect of the present invention, there is provided a syntax encoding method for encoding a bitstream, the method comprising: inputting an image encoding method supporting a spatial random access to a target image, Wherein the code sequentially encodes the blocks in accordance with two or more scan directions having different starting positions so that two or more bit streams are generated for each of the fragment images, A bitstream that is selected from the two or more bitstreams of each piece image having an area overlapping the target image based on the positional information of the target image, The image segment having an area overlapping the target image And the angle is encoded according to the raster scan direction to indicate that it is a generated bit stream.

According to another aspect of the present invention, there is provided a syntax for expressing a structure of a bitstream, the method comprising: providing a video encoding method supporting spatial random access to a target image, Wherein each of the plurality of pieces of image is divided into a plurality of blocks each of which is adjacent to the four vertexes of the rectangle as a start position and the blocks adjacent to the start position are sequentially encoded in a scanning direction And generating four bitstreams for each of the fragmented images by encoding the selected bitstreams in accordance with the positional information of the target image and determining a bitstream selected from among the four bitstreams of each of the fragmented images having an area overlapping the target image, Or the location information of the target image Based on the raster scan direction, each of the piece images having an area overlapping the target image is coded by a raster scan direction to indicate that the bit stream is generated.

According to an aspect of the present invention, there is provided a method of decoding an object image, the object image being a part of a whole image partitioned into a plurality of fragment images by a plurality of boundary lines, The sequential decoding of the adjacent blocks is performed by using the block of the piece image adjacent to the intersection where the boundary line meets as a start position, thereby reconstructing the target image by restoring all or a part of the piece image.

According to an aspect of the present invention, sequential decoding of the adjacent blocks from the block at the start position may be performed based on information indicating the scan direction of the blocks of the piece image.

According to another aspect of the present invention, there is provided a method of decoding a target image, which is a part of a whole image partitioned into a plurality of rectangular piece images by horizontal and vertical boundaries, A block of the rectangular piece image adjacent to an intersection at which the vertical boundary line meets is used as a start position to sequentially decode the adjacent blocks so that all or part of the rectangular piece image is reconstructed to reconstruct the target image.

According to an aspect of the embodiment, when the target image includes the upper right vertex of the rectangular piece image, the block of one row is sequentially encoded from right to left, Or successively encodes a block in a column from the top to the bottom and sequentially encodes the blocks from the top to the bottom of the left column, and the target image is coded in the upper left corner of the rectangular piece image In the case of including a vertex, a method of sequentially encoding a block of a row from left to right and sequentially encoding the blocks from the left to the right of the lower row or a method of encoding a block of a column Blocks are sequentially encoded, and again, And sequentially encodes a block of a row from right to left when the target image includes the lower right corner of the rectangular piece image, and sequentially encodes the block of the right row to the left of the upper row Or successively encodes a block in a column from the bottom to the bottom and sequentially encodes the blocks up to the bottom of the left column, or alternatively, the target image is coded on the left side of the rectangular piece image In the case of including a lower vertex, a method of sequentially encoding a block of a row from left to right and sequentially encoding the blocks from the left to the right of the upper row, And sequentially encodes the blocks of the right column The top block in the future can be sequentially encoded.

According to another aspect of the present invention, there is provided a method of decoding an object image, which is a part of a whole image partitioned into a plurality of rectangular piece images by a horizontal boundary and / or a vertical boundary, Determining a scan direction for the blocks of the rectangular piece image in which all or a part of the target image overlaps with the target image, and sequentially decoding the blocks of the rectangular piece image according to the determined scan direction .

According to an embodiment of the present invention, in the step of performing sequential decoding, decoding is performed using a block of the piece image adjacent to an intersection where the horizontal boundary and the vertical boundary meet.

As described above, according to the image codec system, image encoding method, and image decoding method supporting spatial arbitrary access according to the present invention, the user can view the size of the display window owned by himself / It is possible to receive and decode only a required portion of the entire image. Accordingly, even under a limited transmission channel environment, the user can expect to see only a desired region of the entire image at a high image quality by using his or her own terminal.

Figs. 1A to 1C are diagrams showing the relationship between the size of a piece image and a region to be encoded and transmitted. Fig.
2 is a view showing a raster scanning direction.
3 is a block diagram illustrating a configuration of an image codec system according to an embodiment of the present invention.
4 is a diagram for explaining the relationship between a whole image, a piece image, and a target image, and an example of a coding method in the coding apparatus of FIG.
5A to 5D are diagrams illustrating scan directions of four groups in an encoding apparatus according to an embodiment of the present invention.
6A, 6B, and 6C are diagrams for explaining another example of a coding method in the coding apparatus of FIG.
7A, 7B, 7C, and 7D are diagrams for explaining still another example of a coding method in the coding apparatus of FIG.
8 is a block diagram illustrating a configuration of an image codec system according to another embodiment of the present invention.
Fig. 9 is a diagram for explaining an example of a decoding method in the decoding apparatuses of Figs. 3 and 8. Fig.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. The following embodiments are for the purpose of illustrating the technical idea of the present invention, and therefore the technical idea of the present invention should not be interpreted as being limited by this embodiment. In the description of the following embodiments, the names of the respective components may be referred to as other names in the related art. If they are identical to the functional similarity, even if other names are used, have. In the same way, even if an embodiment in which some of the configurations on the drawings are adopted is adopted, both of them can be regarded as an equivalent configuration if they are identical with the functional similarity. In the description of this embodiment and in the drawings, reference numerals added to respective components are merely described for convenience of explanation.

3 is a block diagram illustrating a structure of an image codec system according to an embodiment of the present invention. Referring to FIG. 3, the video codec system 100 includes a coding server 110 and a decoding apparatus 120. The encoding server 110 includes an encoder unit 112, a memory unit 114, and a bitstream extractor 116. Hereinafter, the configuration and functions of the image codec system 100 will be described in more detail with reference to FIG.

When the entire image is input to the encoder unit 112, the encoder unit 112 first divides the entire image into a plurality of fragment images having a predetermined size. The sizes and the number of the divided and segmented pieces of images can be appropriately selected in consideration of the size of the entire image and the coding efficiency. And the size and the number may be fixed in advance or may be variable according to the characteristics of the entire image. For example, as shown in Fig. 4, the encoder unit 112 divides the input whole image 10 into eight pieces of rectangular image pieces 11 to 18 divided by a horizontal boundary line and a vertical boundary line .

The encoder unit 112 performs encoding on the blocks constituting each of the fragment images 11 to 18, and the encoder unit 112 according to the present embodiment performs the encoding on the blocks of the fragment images 11 to 18 The adjacent blocks are sequentially encoded in different scan directions by different starting positions. For example, when the sculpted images 11 to 18 are in a rectangular shape, the encoder unit 112 performs encoding on each sculptured image 11 to 18 on a block basis, for example, on a macroblock basis. The four kinds of scanning directions refer to coding sequentially for each block image in the piece image, starting from a block corresponding to four starting positions for each piece image 11 to 18. As a result, the encoder unit 112 generates four kinds of bit streams encoded in four kinds of scanning directions in which the start positions are different for each piece image.

Each of the four types of scanning directions may start scanning in a block adjacent to the vertex of each piece image 11 to 18 of a rectangular shape. That is, blocks adjacent to each of the slice images 11 to 18 are coded with the blocks located at the leftmost position, the rightmost position, the leftmost position, and the rightmost bottom position as start positions. Referring to the piece image 11 of FIG. 4, the coding according to the four types of scanning directions includes a block 11a, a block 11b, a block 11c, and a block 11b, which are adjacent to a vertex of the piece image 11 having a rectangular shape. And the block 11d as the start positions, and proceeds coding in the order of the adjacent blocks.

According to an embodiment of the present invention, the four kinds of scan directions may be selected one by one in the group of the following four groups of scan directions. The four groups related to the scanning direction are as follows, and the scanning directions in each group are shown in Figs. 5A to 5D.

(1) As shown in FIG. 5A, the first group is a scanning direction in which a block located at the upper left in the piece image is encoded as a starting position in the direction of a block located at the lower right,

(1) A method of successively encoding a block in a row from left to right and sequentially encoding the blocks from the left to the right of the lower row (left figure in FIG. 5A); And

(2) a method of successively encoding a block of a column from top to bottom and sequentially encoding the blocks from the top to the bottom of the right column (the right picture of FIG. 5A).

(2) The second group, as shown in FIG. 5B, is a scanning direction in which a block located at the upper right in the piece image is encoded as a starting position in the direction of a block located at the lower left,

(1) A method of successively encoding a block of one row from right to left and sequentially encoding the blocks from right to left of the lower row (Fig. 5B, left); And

(2) a method of successively encoding a block in a row from top to bottom and sequentially coding the blocks from the top left to the bottom in the left column (right figure in FIG. 5B).

(3) As shown in FIG. 5C, the third group is a scan direction in which a block positioned at the lower left in the piece image is encoded as a start position in the direction of a block located at the upper right,

(1) A method of successively encoding a block of one row from left to right and sequentially encoding the blocks from the left to the right of the upper row (left figure in FIG. 5C); And

(2) a method of successively encoding a block in a row from the bottom up and then sequentially encoding the blocks from the bottom right to the top in the right column (right figure of FIG. 5C).

(4) As shown in FIG. 5D, the fourth group is a scanning direction in which a block located at the lower right in the piece image is encoded as a starting position in the direction of a block located at the upper left,

(1) A method of successively encoding a block of a line from right to left and sequentially encoding the blocks from the right to the left of the upper row (Fig. 5 (left)); And

(2) a method of sequentially encoding a block in a row from the bottom up and then sequentially encoding the blocks from the bottom row to the bottom row in the left column (right figure in FIG. 5 (d)).

According to another aspect of the present invention, each group of the first to fourth groups may further include a method of sequentially scanning blocks in a zigzag form from the start position of each group.

As described above, in the encoder unit 112, four bit streams generated by performing encoding in accordance with four kinds of scan directions for each piece video 11 through 18 are sent to the memory unit 114, (114) receives and stores the four bit streams for each piece image (11 to 18).

When the location information of a target image desired to be displayed by a user, more specifically, a portable terminal having a decoding apparatus 120 possessed by the user, is received by the image encoding apparatus 110, the bit stream extractor 116 extracts Extracts one bit stream among the four bit streams for each piece image, and transmits the extracted bit stream to the decoding apparatus 120. [ As described above, the process of selecting one bit stream among the four bit streams for each piece image and extracting the bit stream from the memory unit 114 is performed by the bit stream extractor 116. The bitstream extractor 116 selects a bitstream corresponding to a necessary region (an area overlapping the target image) in the fragment image based on the positional information of the target image, or extracts a bitstream that minimizes the amount of data to be transmitted Select. In other words, the bitstream extractor 116 extracts the bitstreams (or the amount of data) transmitted to the decoding device 120 in consideration of the relative direction in which each piece image overlaps with the target image, And the extracted bit stream is transmitted from the image encoding apparatus 110 to the decryption apparatus 120. [0051]

An example of a method of selecting and extracting one bitstream to be transmitted to the decoding apparatus 120 among the four bitstreams for each piece image in the bitstream extractor 116 is as follows. In this case, the information indicating the scanning direction extracted by the bitstream extractor 116 may be included in the bitstream and transmitted to the decoding apparatus 120.

(1) When the target image overlaps with the upper left area of the sculptured image as in the sculptured image 17 of FIG. 4, the bitstream extractor 16 extracts, from among the four bitstreams, Extracted bitstream. For example, in FIG. 4, an area of the target image 20 overlapping the sculptured image 17 is shown to extract a bitstream encoded according to the scan direction shown in the left-hand side of FIG. 5A. Of course, it is also possible to extract and transmit the encoded bit stream according to the scanning direction shown in the right side of FIG. 5A.

2, when the target image overlaps with the upper right area of the sculptured image, the bitstream encoded according to the scanning direction belonging to the second group is extracted from the four bitstreams. For example, in FIG. 4, an area of the target image 20 overlapping the sculptured image 16 is extracted by extracting a bitstream encoded according to the scan direction shown in the left-hand side of FIG. 5B. Of course, the encoded bitstream may be extracted and transmitted according to the scanning direction shown in the right side of FIG. 5B.

3, when the target image overlaps with the lower left area of the sculptured image, a bitstream encoded according to the scan direction belonging to the third group is extracted from the four bitstreams. For example, in FIG. 4, an area of the target image 20 overlapping the sculptured image 13 is extracted by extracting a bitstream encoded according to the scan direction shown in the left-hand side of FIG. 5C. Of course, the encoded bitstream may be extracted and transmitted according to the scanning direction shown in the right side of FIG. 5C.

4, when the target image overlaps with the lower right area of the sculptured image, the bitstream encoded according to the scan direction belonging to the fourth group is extracted. For example, in FIG. 4, an area of the target image 20 overlapping the sculptured image 12 is extracted by extracting a bitstream encoded according to the scan direction shown in the right side of FIG. 5D. Of course, it is also possible to extract and transmit the encoded bitstream according to the scanning direction shown in the left side of FIG. 5D.

FIGS. 6A, 6B, and 6C are diagrams illustrating another example of the above-described embodiment for encoding a target image in the image encoding apparatus 110. FIG.

6A, 6B, and 6C, the encoder unit 112 divides the input whole image 30 into four rectangular pieces 31, 32, and 33 having the same size by the horizontal and vertical boundaries , And 34, and performs encoding on each piece image 31, 32, 33, and 34. FIG. For example, the encoder unit 112 selects one scan direction in each of the four groups for each piece image 31, 32, 33, and 34, and encodes blocks of the piece image according to the selected scan direction To generate four bit streams. Alternatively, the encoder unit 112 selects, for each piece image 31, 32, 33, 34, a scanning direction having a block adjacent to the intersection of the horizontal border and the vertical border as a start position from among the four groups described above And generate one bit stream by performing encoding on the blocks of the fragment image according to the selected scan direction.

Referring to FIG. 6B, four bit streams for each piece image generated in the encoder unit 112 are stored in the memory unit 1114. FIG. According to the present embodiment, the four scan directions selected to be stored in the memory unit 114 correspond to those shown on the left of FIGS. 5A and 5C, respectively, and the second group and the fourth group The groups correspond to those shown on the right side of Figs. 5B and 5D, respectively.

When the location information of the target image 40 (see FIG. 6A) is input from the user to the image encoding apparatus 110, the bitstream extractor 116 of the image encoding apparatus 110 extracts the fragment images One bitstream is selected from the four bitstreams for the first, second, and third bitstreams 31, 32, 33, The bit stream extractor 116 can minimize the amount of transmission data among the four bit streams per piece image in consideration of the positions of the piece images 31, 32, 33, and 34 overlapping the target image 40 And extracts the bit stream encoded in the scanning direction and transmits the extracted bit stream to the image decoding apparatus 120.

6B shows a bit stream according to the scanning direction extracted by the bit stream extractor 116 with respect to each piece image 31, 32, 33, and 34 on the right side of the bit stream extractor 116. FIG. The bitstreams of the fragment images 31, 32, 33, and 34 thus extracted are transmitted to the image decoding apparatus 120.

In this case, not only the entire bitstream of each piece image 31, 32, 33, 34 but also a bit stream as shown by arrows in each piece image 31, 32, 33, 34 of FIG. 6A 32, 33 and 34 overlapping the target image 40, as shown by arrows in Fig. 6C or Fig. 6C.

As described above, according to the present invention, since the encoder unit 112 generates a bit stream by encoding a sculptured image in four scanning directions, and the four bit streams are stored in the memory unit 114, When the user wants a target image to overlap with a part of the piece image, it is possible to extract and transmit only a bit stream corresponding to a necessary part in the piece image, rather than a bit stream for the whole piece image. Therefore, according to the present invention, the amount of data transmitted from the image encoding apparatus 110 to the image decoding apparatus 120 can be reduced.

7A, 7B, 7C, and 7D are diagrams illustrating another example of the above-described embodiment for encoding a target image in the image encoding apparatus 110. FIG.

Referring to FIGS. 7A, 7B, 7C, and 7D, the encoder unit 112 divides the input whole image 50 into four pieces of rectangular images (for example, 51, 52, 53, and 54, and performs encoding on each piece image 51, 52, 53, 54. 7A, the entire image 50 is suitable for an image having a much larger length in the horizontal direction than the vertical direction, for example, a wide image or a panoramic image, and has no horizontal boundary, The entire image 50 is divided into a plurality of pieces of image 51, 52, 53, The height of the target image 60 is less than or equal to the height of the whole image 50 or the piece image 51, 52, 53,

According to the present embodiment, the encoder unit 112 selects one scanning direction in each of the four groups of the above-described embodiments for each piece image 51, 52, 53, and 54 as in the above- And generates four bit streams by performing encoding on the blocks of the piece image according to the selected scan direction. Alternatively, the block 51a adjacent to the upper left vertex and the block 51a adjacent to the upper right vertex, which are shown in the piece image 51 of FIG. 7A, are sequentially encoded with respect to their adjacent blocks as start positions Only two bit streams may be generated. For example, with respect to each piece image 51, 52, 53, and 54, two blocks 51a and 52b of FIG. 7A are coded in two different scanning directions as shown in FIG. A bitstream can be generated.

And two bit streams for each piece image generated in the encoder unit are stored in the memory unit 114. [ When the position information of the target image 60 (see FIG. 7A) is input from the user to the image encoding apparatus 110, the bitstream extractor 116 of the image encoding apparatus 110 extracts the pieces of image data One bitstream is selected from the two bitstreams for the first and second bitstreams 52 and 53, respectively. The bitstream extractor 116 extracts the bitstream of each piece image 52 and 53 from the two bitstreams for each piece image in a scanning direction in which the amount of transmission data can be minimized And transmits the extracted bitstream to the decoding apparatus 120.

On the right side of the bitstream extractor 116 in FIG. 7C, a bitstream according to the scanning direction extracted by the bitstream extractor 116 is shown for each piece image 52, 53. The bit stream of each piece image 52, 53 extracted in this way is transmitted to the decoding device 120. In this case, not only the entire bit stream of each piece image 52, 53 but also a bit stream as shown by an arrow in each piece image 52, 53 of FIG. 7A or 7D may be transmitted.

As described above, according to the present invention, the encoder unit 112 encodes a piece image in two or four scanning directions to generate a bit stream, and the two or four bit streams are stored in the memory unit 114 The bitstream extractor 116 extracts only a bitstream corresponding to a required portion of the fragment image and transmits the extracted bitstream instead of a bitstream of the entire fragment image when the user's desired target image overlaps with a part of the fragment image Do. Therefore, according to the present invention, the amount of data transmitted from the image encoding apparatus 110 to the image decoding apparatus 120 can be reduced.

FIG. 8 shows a structure of an image codec system according to another embodiment of the present invention. Referring to FIG. 8, the video codec system 200 includes a first encoding server 210, a second encoding server 230, and a decoding apparatus 220. The first encoding server 210 includes a first encoder unit 212 and a first memory unit 214 and the second encoding server 230 includes a second encoder unit 232 and a second memory unit 234 ), And a bitstream extractor 236.

The first encoding server 210 performs the same operations and functions as the encoding server according to the related art. For example, the first encoding server 210 may be a device for encoding a target image in real time by a user. More specifically, when the position information of the target image is received from the user's decoding apparatus 220, the first encoder unit 212 encodes the target image inputted in the entire image in real time and outputs the encoded data to the first memory unit 214 and transmits the bitstream of the input target image to the decoding device 220. [

The second encoding server 230 is an encoding server according to the above-described embodiment of the present invention. That is, the second encoder unit 232 divides the entire image into a plurality of fragmented images having a predetermined size, generates four bitstreams encoded in four scan directions for each fragmented image, (234). The bitstream extractor 236 selects and extracts one bitstream from the four bitstreams for each piece image based on the location information of the target image received from the decoding device 220. The bit stream of each piece image thus selected is transmitted from the second encoding server 230 to the decoding apparatus 220.

Therefore, the encoding servers 210 and 230 of the image codec system 200 according to the present embodiment perform encoding in one of the encoding servers 210 and 230 according to a user's request, and generate bitstreams To the decryption apparatus 220. From the viewpoint of the decoding apparatus 220, it is possible to recognize the bit stream transmitted from the encoding server 210 or 230 first by using a certain encoding server so that the target image can be correctly reconstructed.

As a result, according to the present embodiment, it is determined whether the first encoding server 210 has encoded the target image in real time according to the conventional technique, or one of the embodiments of the present invention described in the second encoding server 230 It is necessary to generate information for identifying whether the encoding has been performed according to the bitstream and to include the information in the bitstream transmitted to the decoding apparatus 220. [ Since this information is used to indicate that one of the two encoding servers 210 and 230 has been encoded by the encoding server, it can be expressed using 1-bit information. That is, one bit indicating the ON / OFF mode for any one of the encoding servers may be added to the bit stream transmitted to the decoding apparatus 220. When the 1-bit information added to the bit stream for compatibility between the first encoding server 210 and the second encoding server 230 is a bit stream encoded by the first encoding server 210, for example, Quot; 1 "in the case of the bit stream coded by the second encoding server 230, and vice versa.

Next, a decoding method of the decoding apparatuses 120 and 220 of the image codec systems 100 and 200 according to the embodiment of the present invention will be described with reference to FIG. Here, the drawing of Fig. 9 corresponds to the drawing of Fig. 6A. In the case of the decoding apparatus 220 of the image codec system 200, in the case of decoding the input bit stream encoded by the second encoding server 230, that is, when the second encoding server 230 encodes Only one bit of information indicating that the input bit stream is included is applied. Therefore, when decoding the input bitstream encoded by the first encoding server 210 of the video codec system 200, a decoding method described later is not applied and the same decoding method as the conventional method is applied.

The decoding method described below is not limited to the case of the decoding apparatuses 120 and 220 in the video codec systems 100 and 200 described above, It can also be applied to decoding a stream. That is, a decoding method described below can be applied, in which a whole image can be divided into a plurality of fragment images, and a bitstream encoded according to different scan directions is decoded for each fragment image.

9, a decoding method in the decoding apparatus 120, 220 according to the present embodiment is a decoding method in which a target image 40 is divided into a piece image 31 , 32, 33, and 34) in the different scanning directions.

Herein, the block adjacent to the one point means that when the entire image is divided into a plurality of pieces of images by one or a plurality of horizontal boundaries and one or a plurality of vertical boundaries, the block adjacent to one intersection where the horizontal boundary and the vertical boundary meet to be. For example, when the entire image 30 is divided into four piece images 31, 32, 33, and 34 by one horizontal boundary line 52 and one vertical boundary line 54 as in the example shown in FIG. 9 The block adjacent to the one point is a block 31a, 32a, 33a, 34a adjacent to the intersection 56 of the horizontal boundary line 52 and the vertical boundary line 54. The block adjacent to the one point may be a block adjacent to all intersections where the plurality of boundary lines meet when the whole image is divided into a plurality of pieces of images such as a triangle or a rhombus other than a rectangle by a plurality of boundary lines.

According to the decoding method according to the embodiment of the present invention, the area of the target image 40 overlapping the sculpted image 31 (the sculptured image positioned on the upper left with respect to the intersection) And the area of the target image 40, which overlaps with the piece image 32 (the piece image positioned at the upper right with respect to the intersection), is sequentially decoded in the upper right direction with the block 32a as the start position And the area of the target image 40 overlapping with the sculptured image 33 (the sculpture image positioned at the lower left with respect to the intersection) sequentially decodes in the lower left direction with the block 33a as the start position, and the sculptured image 34 (The piece image positioned at the lower right with respect to the intersection) is overlapped with the start position of the block 34a And it performs decoding to the lower right direction by one.

The reason for performing the decoding in this manner is that the decoding apparatuses 120 and 220 must decode the input bitstream in the scanning direction encoded by the encoding servers 110 and 210. Therefore, in the case of the area of the target image 40 overlapping the sculptured image 31, decoding is performed in the left direction from the block 31a at the start position, and then decoding is performed in the left direction in the upper row, (I.e., decoding in accordance with one scanning direction among the scanning directions belonging to the fourth group) after performing the decoding in the upward direction from the first group 31a. In this way, in the case of the area of the target image 40 overlapping the sculptured image 31, information indicating which scan direction to perform decoding may be included in the input bitstream.

This principle can also be applied to the case of the region of the target image 40 overlapping with the other piece images 32, 33, For example, in the case of the area of the target image 40 overlapping with the sculptured image 32, decoding is performed according to one of the scan directions belonging to the third group, and the area of the target image 40 overlapping the sculptured image 33 And in the case of the region of the target image 40 overlapping the sculptured image 34, one of the scan directions belonging to the first group is decoded, The decoding can be performed according to the direction.

As described above, in the case where a block adjacent to the intersection of the horizontal boundary line and the vertical boundary line for dividing the entire image into each piece image is used as a starting point, and each piece image in the outer direction (horizontal priority direction, vertical priority direction, or zigzag direction) It is possible to reconstruct the entirety of the target image that the user wants to display by restoring the images corresponding to the regions of the target image overlapping the target image.

Claims (8)

A method of encoding an input image input to a video encoding apparatus,
Determining the number and size of the piece images constituting the input image;
Dividing the input image into a plurality of fragment images according to the determined number and size;
An encoding step of performing parallel encoding on each of the plurality of fragment images; And
Generating partition information, which is information indicating whether the input image is divided into a plurality of fragment images,
The segment information is represented by a 1-bit flag for each input image,
Wherein the input image is divided into a plurality of pieces of the piece image having a rectangular shape along a horizontal boundary line,
Wherein each of the plurality of piece images comprises a plurality of blocks,
Wherein each of the plurality of blocks is included only in one piece image. The method according to claim 1,
Wherein the encoding step sequentially encodes blocks in the piece image using one of a first scan direction, a second scan direction, a third scan direction, and a fourth scan direction,
Wherein the first scanning direction is a scanning direction in which a left-upper corner block in the piece image is sequentially encoded in a horizontal or vertical direction with a start position,
Wherein the second scanning direction is a scanning direction in which a right-upper corner block in the piece image is sequentially encoded in a horizontal or vertical direction with a start position,
Wherein the third scanning direction is a scanning direction in which a left-lower corner block in the piece image is sequentially encoded in a horizontal or vertical direction with a start position,
Wherein the fourth scan direction is a scan direction for sequentially encoding a horizontal or vertical direction with a right-bottom corner block in the piece image as a start position. The method according to claim 1,
Wherein each piece image is a rectangular shape. delete A method for decoding a coded input image generated by coding an input image,
Receiving partition information, which is information indicating whether the input image is divided into a plurality of fragment images, in the process of coding the input image;
Obtaining information on the number and size of the piece images constituting the input image when the segment information indicates that the input image is segmented into a plurality of piece images;
Dividing the encoded input image into the plurality of fragment images based on the information on the obtained number and size; And
And performing parallel decoding on each of the plurality of fragment images,
The segment information is represented by a 1-bit flag for each input image,
Wherein the encoded input image is divided into a plurality of pieces of piece images each having a rectangular shape along a horizontal boundary line, each of the plurality of piece images is composed of a plurality of blocks, and each of the plurality of blocks includes an image Decoding method. 6. The method of claim 5,
Wherein the decoding step sequentially decodes blocks in the piece image using one of a first scanning direction, a second scanning direction, a third scanning direction, and a fourth scanning direction,
Wherein the first scanning direction is a scanning direction for sequentially deciphering a horizontal or vertical direction with a left-upper corner block in the piece image as a start position,
Wherein the second scanning direction is a scanning direction in which a right-upper corner block in the piece image is decoded in a horizontal or vertical direction with a start position,
Wherein the third scanning direction is a scanning direction in which a left-lower corner block in the piece image is decoded in a horizontal or vertical direction with a start position,
Wherein the fourth scan direction is a scan direction in which a horizontal or vertical direction is sequentially decoded with a right-bottom corner block in the piece image as a start position. 6. The method of claim 5,
Wherein each piece image is a rectangular shape. delete

Images