TWI508530B - Image compression methods, media data files, and decompression methods - Google Patents

Description

Image compression method, media data file and decompression method

The invention relates to an image compression method, a related media data file and a decompression method, and more particularly to an image compression method which can moderately reduce the amount of data and maintain image quality.

In mobile communication devices, audio-visual entertainment is often an indispensable part. For example, being able to click and play a movie on a mobile phone or tablet anytime and anywhere is an application that most users want or are used to.

Generally speaking, the media data file of a movie is to present a sufficiently delicate image on the large screen of a television or a projector. Therefore, the resolution of each frame in the media data file will be quite high, with the media. The data file will be quite large. Such a file to be played in a mobile communication device has the following disadvantages.

1. The picture may be delayed. A huge media data file requires a high-speed communication network to maintain smooth video playback. However, the communication network bandwidth of the mobile communication device is often limited, so it is easy to cause an event of the screen stop delay.

2. Need more storage hardware. Storage hardware in mobile communication devices is more expensive than conventional devices. To play high-resolution images, more storage media is needed, which is quite extravagant for mobile communication devices.

3. Shorten the operating time of the mobile communication device. Processing high-resolution images requires more computation, and therefore consumes more power, thus affecting the operating time of the mobile communication device.

To this end, how to provide an effective image compression method, in addition to reducing the size of the media data file, while properly retaining the quality of each frame in the media data file, is very much needed.

An embodiment of the present invention provides an image compression method, including: dividing a original image frame into a first portion and a second portion; reducing the second portion to generate a reduced portion; and, reconstructing The first portion and the reduced portion are used to generate a reconstructed frame. The reconstructed picture frame has the same picture frame size as the original picture frame.

An embodiment of the present invention provides a method for decompressing a media data file. The decompression method includes: generating a reconstructed image frame and an auxiliary information from the media data file; and finding one of the reconstructed image frames according to the auxiliary information. a first portion and a reduced portion; the reduced portion is enlarged to generate a blurred portion; and the first portion and the blurred portion are reconstructed according to the auxiliary information to generate a combined image frame.

Embodiments of the present invention provide a media data archive that conforms to a particular file format. The media data file contains a plurality of first objects and a plurality of second objects. The first object includes media material having a plurality of reconstructed frame frames. Each of the second objects includes an auxiliary information corresponding to the reconstructed frame and auxiliary information. The affiliate information can be used to decompress the media data to generate the corresponding reconstructed map frame. The auxiliary information can be used to find a first part and a reduced part of the corresponding reconstructed picture frame, and record a reduction ratio of the reduced part.

Embodiments of the present invention provide a media data file that includes a video file and an interpretation data. The video file conforms to a specific file format, and after decoding, a plurality of reconstructed frame frames and corresponding audio information are obtained. The interpretation data contains auxiliary information corresponding to the reconstructed frame. The auxiliary information can be used to find a first part and a reduced part of a corresponding reconstructed picture frame, and record a reduction ratio of the reduced part.

In order to better understand the above and other aspects of the present invention, the preferred embodiments are described below, and in conjunction with the drawings, the detailed description is as follows:

In this specification, elements or signals of the same symbol are elements or signals having the same or similar functions. Those of ordinary skill in the art can deduce the elements or signals of the same symbols in accordance with the teachings of the present specification, and the embodiments thereof may not be limited to the one disclosed in the present specification, but may have many different variations.

FIG. 1 shows an image compression method 10 implemented in accordance with the present invention, which is applicable to a media data file and can be used in an encoder. The media data file has several original image frames for playing in sequence. After each original picture frame is processed by the image compression method 10, a corresponding reconstructed picture frame is generated. The reconstructed picture frame is composed of the interested part, the uninterested part and the auxiliary information in the original picture frame. In this embodiment, the interested portion is defined as the first weight portion, and the non-interest portion is the second weight portion, wherein the first weight is higher than the second weight, and regarding the weight selection, for example, the user's demand lies in the subtitle Clear, the weight of the subtitle part can be increased. If the user's needs are the clarity of a specific object in the frame (such as a face, a human figure or an object, etc.), the weight of the corresponding part can be increased. For example, a certain proportion of the lower part of the picture, or a white or high contrast part of the picture can be used as a reference for judging the part as a subtitle.

Alternatively, defining a lower portion of the screen at a certain proportion, the color is close to white and the same color block extending vertically, horizontally or at an angle is a stroke area, and the stroke area and a certain range around it are defined as the first weight portion. In this embodiment, after determining the weight of the weight, the reconstruction of the frame is performed, and the pixels in the second weight portion with lower weight are correspondingly reduced to the lower resolution image, where the reduction includes Scale down and non-equal reduction. The first weight portion with a higher weight is not subjected to the reduction operation and is placed in the original position in the frame where the frame is not processed. The reduced second weight portion is rearranged in the frame, because the second weight portion is reduced, and after processing, the frame will generate a blank that does not belong to the first weight portion or the second weight portion. This part can be filled with black/white as the background color to reduce the amount of information. It can also be filled in with other color levels, and the purpose of reducing the amount of information can also be achieved. In other words, the resolution of the portion of interest of the user remains unchanged, and the resolution of the uninterested portion is reduced (relative to the original frame). Therefore, the new media data file generated by the compression coding of the reconstructed image frame can be smaller than the original media data file to achieve the purpose of reducing the file size, and at the same time, the interested part of the original picture frame can be completely preserved. In addition, the first weight portion and the second weight portion may be compression-encoded at different compression rates, respectively, as long as the compression ratio of the portion of interest is lower than the compression ratio of the non-interest portion, so that the restored portion of the interest is resolved. The degree is higher than the resolution of the uninterested part.

In step 12, the image compression method 10 receives an original picture frame from a media data file. Figure 2 illustrates an original picture frame 30 that will be used to explain the results of some of the steps in the image compression method 10.

Step 14 divides the original map frame 30 to produce at least one interesting portion and at least one non-interest portion that do not overlap. The so-called interest and non-interest parts are relative here. To put it simply, the part of interest is probably the part of the picture quality that you don't want to be sacrificed; the part that is not interested is probably the part of the picture quality that can be sacrificed. For example, each original picture frame can be defined as a plurality of image blocks arranged in a matrix; each image block is roughly a square, and is 16×16 (or 8×8) image pixels (image pixel) It is composed of; each image pixel has a number of pixels corresponding to the three primary colors of red, green and blue. Step 14 can be checked one by one. If a currently checked image block conforms to a preset rule, then the image block can be subordinate to the interested portion; otherwise, the image block is attached to the uninterested portion. This preset rule can be arbitrarily defined by the user. For example, in an embodiment, as long as the image block is located within a quarter of the original frame frame, because it generally contains the caption message, all belong to the interested part; other image blocks are Belongs to the non-interest part. In another embodiment, if one of the image pixels in an image block has a comparison relationship exceeding a certain degree, the image block belongs to the interested portion; otherwise, the image block belongs to the non-interest portion. In another embodiment, if an image block has a stroke portion, the image block belongs to the interested portion; otherwise, the image block belongs to the non-interest portion.

Figure 3 shows some of the possible outcomes of the original image frame 30 after processing through step 14. Step 14 may generate a raw location information map 38. In the original location information map 38, the marked area corresponds to the portion of the original picture frame 30 where the sun and the subtitle are located. Since the contrast is large or contains a stroke, in this embodiment, the part of interest is attached. The blank area in the original location information map 38 corresponds to the non-interest part. Therefore, the original picture frame 30 is divided into interested parts 32 and 34, and no interest part 36. In Fig. 3, frame 33 shows a frame composed of interested portions 32 and 34.

Next, at a reduced scale, step 16 scales down the non-interest portion 36 to produce a reduced portion 36a, as shown in FIG. It can be inferred that the resolution of the reduced portion 36a will be lower than the non-interest portion 36.

Step 18 recomposes the reduced portion 36a and the interested portions 32 and 34 to form a reconstruct frame having the same frame size as the original frame 30. Figure 5A illustrates a reconstructed map frame 40 that maintains the original position and size of the portions of interest 32 and 34, while the reduced portion 36a is located in an area of the reconstructed map frame 40 in which the portions of interest 32 and 34 are not occupied.

In one embodiment, step 18 is to first copy and place the interested portions 32 and 34 in a blank reconstructed map frame, and keep the relative positions and sizes of the interested portions 32 and 34 unchanged. Then, according to a rule, it is determined that the reduced portion 36a is located substantially at one of the placement positions in the reconstructed map frame. Based on the determined placement position, the reduced portion 36a is placed in a blank area in which the reconstructed map frame has not been occupied to complete the reconstructed map frame 40. The rules for determining the placement location can be tailored to the user's preferences.

For example, the reduced portion 36a may attempt to place all of the possible placement positions of the reconstructed map frame, with some of the placement locations causing the placed reduced portion 36a to partially overlap the portion of interest 32 or 34. If a particular placement position is such that the reduced portion 36a being placed does not overlap the interest portions 32 and 34 at all, or the overlap portion is minimized, then the reduced portion 36a is placed at the particular placement position to produce the final reconstructed map. frame. As illustrated in FIG. 5A, the reconstructed map frame 40 is completed, wherein the reduced portion 36a does not overlap the interesting portion 32 or 34.

In another embodiment, the determined placement position may be such that, relative to the reconstructed frame, the size of the generated frame data is minimized after compression according to the compression standard of MPEG-4. Each of the possible placement positions of the reduced portion 36a produces a corresponding reconstructed frame. After compression by the MPEG-4 compression standard, a corresponding frame data is generated, and a corresponding data size is generated. Therefore, as long as the minimum data size is found, the corresponding placement position can be used as the determined placement position.

In the process of generating the reconstructed map frame 40 of Figure 5A, step 18 also generates side information. FIG. 5B illustrates auxiliary information 37 including an original location information map 38 and reconstruction information 42. As previously described, the home location information map 38 contains the original location information of the interested portion 32 or 34 and the non-interest portion 36. The reconstruction information 42 then includes a reduction ratio of the reduced portion 36a and a placement position of the reduced portion 36a in the reconstructed map frame 40. The auxiliary information 37 is not limited to the contents exemplified in FIG. 5B, but may include information desired by other users. Regarding the implementation of the reconstruction information 42 of the interested portion, when compressing using the MPEG format, a weight may be added to the auxiliary or custom region in the compression unit (for example, an 8x8 pixel or a 16x16 pixel unit) including the interested portion. A parameter to define whether the compression unit belongs to an interested part, and the weight parameter may be a two-digit 0 or a 1 value, and the weight parameter of the interested portion is defined as 1. When there is overlapping data in the reconstructed picture frame, the part with the weight parameter of 1 is moved to the blank part of the non-interest part or the non-interest part 36, and is pre-defined (for example, from left to right, from top) The way to the bottom is arranged along the blanks. In this way, when the map frame is restored, when the weight parameter is 1, but there is no position of the pixel data, the pixel data is restored to the original position of the interested portion according to the predefined arrangement.

Although in this embodiment, the portion of interest is placed at the original position of the frame during reconstruction, it is not limited thereto, and the portion of interest may be placed in other positions of the frame during reconstruction. on. Taking MPEG as an example, after the data is converted to the frequency domain, the high-frequency components of the continuously arranged data are large, and a large amount of compression can be obtained when compressing. Therefore, the data should be continuously arranged as a principle in the reconstruction. The image block is placed such that the reconstructed picture frame has the greatest compression ratio when compressed. When the location of the portion of interest is also re-adjusted under reconstruction, the reconstruction information 42 further includes location information of the portion of the reconstructed map.

Once the placement position of the reduced portion 36a in a reconstructed frame is determined, it is inevitable that the reduced portion 36a may partially overlap the interested portion 32 or 34. To solve this problem when overlapping events occur, first use the appropriate scale reduction parameters. For example, when the width portion of the interested portion occupies one-third of the frame, just select the appropriate reduction ratio to reduce the size. The width of the post-reduced portion is less than two-thirds of the frame of the picture, and the problem of overlap can be solved. However, in the case where the user's demand is that the subtitles are clear and the white or high contrast color blocks in the picture are used as the judging subtitles, other high contrast or white parts may be judged as high-weight reserved parts, which makes them interested. Partially present an irregular situation. In this case, the reduced portion of the block is a complementary shape of the block of interest, and the two may be interlaced with each other. In other words, the same height and width may include the partial block of interest. And a reduction. As a result, it is not possible to simply use the width or height adjustment to avoid the problem of overlapping between the interested portion and the reduced portion. At this time, the portion of the reduced portion 36a or one of the interested portions 32 and 34 may be placed in the reconstructed map frame according to a predetermined shifting manner and rules, and the portion 36a and the interested portion are not yet reduced. The area occupied by 32 and 34. Fig. 6 illustrates another reconstructed map frame 40a in which the placement position of the reduced portion 36a is approximately the upper left corner of the reconstructed map frame 40a, and thus the reduced portion 36a partially overlaps the interested portion 32. The image blocks 44a to 44d display portions of the reduced portion 36a that overlap with the portion of interest 32. The image blocks 44a to 44d are not placed in the region where the interesting portion 32 is located in the reconstructed image frame 40a, but are placed in the regions 46a to 46d according to a predetermined moving manner and rules. In the reconstructed map frame 40a, the overlapped portion in the reduced portion 36a is moved. In another embodiment, the overlapped portion of the portion of interest is moved, and the predetermined shifting manner is based on the order from top to bottom from left to right.

Step 20 encodes the reconstructed picture frame generated in step 18 to generate picture frame data. For example, according to the compression standard of MPEG-4 or other image coding protocol, the reconstructed picture frame is encoded to generate corresponding picture frame data.

Step 22 combines the frame data with the auxiliary information to generate a media data file. In an embodiment, the plurality of frame data is stored in an MP4 file, and the auxiliary information corresponding to the frame data is stored in a metadata file, and the media data file generated in step 22 is an MP4 file and an interpretation data. The combination of files. In another embodiment, the media data file generated in step 22 is only an MP4 file, and the auxiliary information is stored in a field that the user can make in the MP4 file.

Compared to the non-interest portion 36 in the original picture frame 30, the resolution of the reduced portion 36a in the reconstructed picture frame 40 is relatively low, and the reconstructed picture frame 40 also has a plurality of blank portions. Therefore, it can be expected that the media data file generated according to the reconstructed frame should be relatively small, and is suitable for mobile communication devices.

The media data file generated in step 22 can be transmitted to a mobile communication device via a wired or wireless network. As long as the mobile communication device has a corresponding decoding program or decoder, the combined frame frame which is very close to the original image frame can be generated and displayed according to the frame data and the auxiliary information in the media data file.

Figure 7 shows a decompression method 60 that can be applied to a decoder for processing the media data files generated in Figure 1. The decompression method 60 is basically the inverse of the image compression method 10 of FIG.

Step 62 receives a media data file. In one embodiment, the decompression method 60 is used in a mobile phone that receives media data files over a wireless network.

According to a decoding protocol, step 64 decodes the frame data in the media data file to generate a reconstructed frame. For example, if the frame data is compressed and encoded according to the compression standard of MPEG-4, then step 64 is based on the decompression standard of MPEG-4 to roughly restore the reconstructed frame. Because compression and decompression have been experienced, the reconstructed map frame produced by step 64 is substantially identical, even though it is not identical to the reconstructed map frame generated in step 18 of Figure 1. Step 64 also finds the corresponding auxiliary information from the media data file.

Based on the original location information map 38 and the reconstruction information 42 in the auxiliary information, step 66 finds the interested portion and the reduced portion from the reconstructed map frame generated in step 64. For example, according to the original position information map 38 in FIG. 5B, the interested portions 32 and 34 can be found from the reconstructed map frame 40 of FIG. 5A. According to the reconstruction information 42 in Fig. 5B, the reduced portion 36a can be found from the reconstructed map frame 40 of Fig. 5A.

Similarly, the original location information map 38 and the reconstruction information 42 may also know whether the reconstructed map frame 40 has an overlapping portion. As long as the image compression method 10 is known, for predetermined shifting methods and rules for overlapping events, step 66 may find and/or piece together the interested portions 32 and 34 and the reduced portion 36a from the reconstructed map frame 40.

Step 68 amplifies the reduced portion 36a to form a blurred portion having the same size as the reconstructed frame. Taking the reduced portion 36a of Fig. 4 as an example, since the reduction and enlargement process has been performed, the blurred portion produced in step 68 is substantially the same as the non-interest portion 36, but has a lower resolution.

Based on the original location information map 38, step 70 reconstructs the blurred portion and the portion of interest to form a combined map frame. The blurred portion is placed at the position corresponding to the uninterested portion. In general, the combined picture frame and the corresponding original picture frame will have the same interested part, but the blurred part in the combined picture frame will be more fuzzy than the uninterested part in the corresponding original picture frame. The intersection of the blurred part and the interested part can be slightly processed to reduce or avoid image discontinuity caused by the difference in resolution. The combined frame generated by step 70 is then played back in step 74.

Although the resolution of the blurred part is lower in the combination frame, because the information is inherently less important or uninteresting, it may not be visible to the naked eye in the small screen on the mobile phone or tablet, so it is blurred. Part of it is acceptable. As for the interested parts that the user may care about, the combined frame still maintains its original resolution. Therefore, the original picture frame is replaced by the combined picture frame to play, which does not substantially affect the user's feeling of viewing.

In an embodiment of the present invention, the media data file generated by the image compression method 10 of FIG. 1 is an MP4 file, which conforms to the file format of the MP4. The MP4 file contains several objects, each of which is called an atom. For example, the real media material for the reconstructed map frame 40 of Figure 5A is stored in a media data atom. Media data atoms are generally referred to as mdat atoms. The auxiliary information such as the compression mode, the track type, and the time stamp of the reconstructed picture frame 40 is stored in a movie atom, which is generally called a moov atom. The auxiliary information 37 corresponding to the reconstructed map frame 40 is stored in a user data atom that can be customized by the user in the moov atom. Therefore, in an embodiment of the present invention, a movie atom stores an auxiliary information corresponding to a reconstructed frame and auxiliary information. Alternatively, auxiliary information and affiliate information can be appended to the MPEG material. In another embodiment, when the reconstructed frame 40 is compressed by the H.264 method, the auxiliary information and the auxiliary information may be appended to each transmitted frame (for example, frame 1 + first auxiliary information, Figure frame 2+ second auxiliary information...). The auxiliary information is important in the present invention, and the auxiliary information is lost. The reconstructed image frame 40 can only be restored by using the decompression method, and cannot be restored to the original frame frame. Therefore, the above several different auxiliary information are attached. The method is not limited to use alone, and the auxiliary information may be attached to the different positions at the same time to avoid losing the auxiliary information when the frame is lost.

In another embodiment of the present invention, the media data file generated by the image compression method 10 of FIG. 1 is a combination of an MP4 file and a metadata. This MP4 file stores the real media data of all reconstructed frames and the corresponding affiliate information. That is, after directly decompressing the MP4 file with the MPEG-4 decompression standard, several reconstructed picture frames and corresponding audio information can be obtained, but the corresponding auxiliary information cannot be obtained. The interpretation data stores auxiliary information and time stamps corresponding to all reconstructed image frames. The timestamp allows the decoder to quickly find the corresponding auxiliary information when processing a reconstructed frame.

In the embodiment of Fig. 1, the image compression method 10 divides only the original picture frame into two parts: an interesting and a non-interest part. However, the present invention can also divide the original picture frame into two or more parts. For example, in another embodiment of the invention, the original picture frame is divided into three parts: interested, uninterested, and very uninteresting parts. Non-interested and very uninteresting parts can be reduced with different reduction ratios, and then reconstructed with a portion of interest to reconstruct a reconstructed frame. At this time, the auxiliary information may include original location information maps with interest, no interest, and very uninteresting parts, reduction ratios and placement positions of non-interest parts, reduction ratios and placement positions of very uninteresting parts, etc. .

In conclusion, the present invention has been disclosed in the above preferred embodiments, and is not intended to limit the present invention. A person skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention. Therefore, the scope of the invention is defined by the scope of the appended claims.

10. . . Image compression method

12, 14, 16, 18, 20, 22. . . step

30. . . Original frame

32, 34. . . Interested part

33. . . Frame

36. . . Uninterested part

36a. . . Reduced part

37. . . Auxiliary information

38. . . Original location map

40, 40a. . . Reconstructed frame

42. . . Refactoring information

44a~44d. . . Image block

46a~46d. . . region

60. . . Decompression method

62, 64, 66, 68, 70, 74. . . step

Figure 1 shows an image compression method implemented in accordance with the present invention.

Figure 2 illustrates an original picture frame.

Figure 3 shows some of the possible outcomes of a raw frame after processing through step 14.

Figure 4 shows an inconspicuous portion reduced to a reduced portion.

Figure 5A illustrates a reconstructed picture frame.

Figure 5B illustrates the auxiliary information.

Figure 6 illustrates another reconstructed picture frame.

Figure 7 shows a decompression method.

12, 14, 16, 18, 20, 22. . . step

Claims (17)

An image compression method includes: dividing a original frame into a first portion and a second portion; and reducing the second portion to generate a reduced portion; reconstructing the first portion And the reduced portion to generate a reconstructed frame; and encoding the reconstructed frame to generate a frame data; wherein the reconstructing step comprises: determining, according to a rule, the reduced portion is located in the reconstructed map One of the frames is placed in position and the rule is related to the size of the frame data. An image compression method includes: dividing a original frame into a first portion and a second portion; and reducing the second portion to generate a reduced portion; reconstructing the first portion And the reduced portion to generate a reconstructed frame; and encoding the reconstructed frame to generate a frame data; wherein the reconstructing step comprises: determining that the reduced portion is placed in one of the reconstructed frames a position; and when the placement position overlaps the reduced portion with the first portion, placing the reduced portion or one of the first portion overlaps the reconstructed frame in a predefined manner The area that has not been occupied by the first part and the reduced part. An image compression method includes: dividing a original frame into a first portion and a second portion; and reducing the second portion to generate a reduced portion; reconstructing the first portion And the reduced portion to generate a reconstructed frame; and encoding the reconstructed frame to generate a frame data; wherein the reconstructing step comprises: placing the first portion in the reconstructed frame; And placing the reduced portion in an area of the reconstructed frame that is not occupied by the first portion. An image compression method includes: dividing a original frame into a first portion and a second portion; and reducing the second portion to generate a reduced portion; reconstructing the first portion And the reduced portion to generate a reconstructed frame, wherein the reconstructing step determines, according to a rule, that the reduced portion is located at a position of the reconstructed frame, and the rule is related to the reduced portion and the first One of the overlapping areas; and encoding the reconstructed frame to generate a frame of data. The image compression method of claim 4, wherein the original image frame comprises a plurality of image blocks of the same size, each image block has a plurality of pixels, and the segmentation step is based on a A rule is set to determine whether a corresponding image block belongs to the first part or the second part. The image compression method of claim 5, wherein the preset rule is a comparison relationship in the corresponding image block. The image compression method of claim 5, wherein the preset rule is whether the corresponding image block has a stroke portion. The image compression method of claim 4, further comprising: encoding the reconstructed frame to generate the frame data, and generating a media data file, wherein the frame data is included The auxiliary information includes a placement position indicating the reduced portion in the reconstructed frame. The image compression method of claim 8, wherein the auxiliary information includes a reduction ratio indicating one of the reduced portions. The image compression method of claim 8, wherein the auxiliary information includes original information of the first portion and the second portion. A storage medium for storing a media data file, the media data file conforming to a specific file format, and comprising: a plurality of first objects, the media data comprising a plurality of reconstructed frames; a plurality of second objects, each of the second objects includes auxiliary information of a corresponding reconstructed frame in the reconstructed frame and auxiliary information; wherein the auxiliary information is used to decompress the media data to generate the corresponding reconstructed image And the auxiliary information may be used to find a first part and a reduced part of the corresponding reconstructed picture frame, and record a reduction ratio of the reduced part, and the auxiliary information includes at least the first Part of the original location information. The storage medium of claim 11, wherein the specific file format is an MP4 file format, and the auxiliary information is stored in a movie object, and a user can customize a user data item. A storage medium for storing a media data file, the media data file comprising: a video file, which conforms to a specific file format, and is decoded to obtain a plurality of reconstructed frames and corresponding audio; and the interpretation data includes There are auxiliary information corresponding to the reconstruction frame and a time stamp; wherein the auxiliary information can be used to find a first part and a reduced part of a corresponding reconstructed picture frame, and the recording has one of the reduced parts The ratio is reduced, and the auxiliary information includes at least the original location information of the first part. The storage medium of claim 13, wherein the specific file format is an MP4 file format. A method for decompressing a media data file, the decompression method comprising: generating a reconstructed image frame and an auxiliary information from the media data file; and finding a first part and a first part of the reconstructed image frame according to the auxiliary information Reducing the portion; enlarging the reduced portion to generate a blurred portion; and reconstructing the first portion and the blurred portion to generate a combined image frame according to the auxiliary information; wherein the auxiliary information includes the The first part and the original position information of the blurred part. The decompression method according to claim 15, wherein the auxiliary information includes a reduction ratio, and the amplification step generates the blurred portion according to the reduction ratio. The decompression method of claim 15, wherein the reconstructed frame comprises a blank portion.