CN102868869B - Image compression method and decompression method with fixed compression ratio and electronic device thereof - Google Patents

Abstract

The invention discloses an image compression method, a decompression method and an electronic device with fixed compression ratio. The encoder compresses the original data of the image into an image bitstream at a compression rate by the following steps: dividing the image into a plurality of image rows; appointing a storage space in a memory according to the compression ratio and the bit stream length of the original data; dividing the storage space into a plurality of compression arrangement areas according to the compression multiplying power and the bit stream length of the original data; recording the initial position of the compressed row area as the current termination position; receiving the chart rows in sequence; compressing the image sequences into compressed bit streams respectively; and sequentially writing the compressed bit stream into the corresponding compressed row area according to the group consisting of the field index of the picture row, the output field sequence and the image row corresponding to the picture row.

Description

Image compression method and decompression method with fixed compression ratio and electronic device thereof

technical field

The present invention relates to an image compression method, a decompression method and an electronic device thereof, in particular to an image compression method, a decompression method and an electronic device thereof with a fixed compression ratio for digital image raw data.

Background technique

Digital cameras and other devices initially obtain data from multi-field sensors such as charge coupled devices (CCDs) as raw data, which must be processed by image processing devices. A lot of processing can be provided for users to watch. However, the existing technologies for compressing raw data mostly use lossless compression or non-visual lossy compression, and it is difficult to control the compressed file size. The space saved by this kind of compression method is not certain, and it can only be confirmed how much storage space is saved after the actual compression is completed. The latest image signal processor (image signal processor, ISP) mostly uses image band as the unit for processing. However, the multi-field sensor does not output the original data in a fixed order, so it is difficult to simultaneously consider the field order output by the multi-field sensor and the processing in units of image rows. Especially in the compressed bit stream, it is very difficult to balance the above two to define the boundary of the image field and the video row.

In addition, even if only part of the image is needed, the most intuitive way is to decode the entire compressed result back to full resolution (full resolution), and then select the required part for further processing. This approach requires the provision of memory capable of holding full-resolution images. And with the progress of the times, the resolution of images increases rapidly, which increases the burden of decoding.

Contents of the invention

In order to solve the above problems, the object of the present invention is to provide an image compression method with a fixed compression ratio, an image decompression method and an electronic device thereof, which can solve the problem that it is difficult to take into account the image field order and the image row boundary in the prior art.

The electronic device includes a multi-field sensor, a memory and a codec wrapper module. The multi-field sensor generates raw data of an image in a field order, wherein the image includes multiple lines, and each line corresponds to a field index. The codec wrapper module includes at least one encoder and at least one decoder, which respectively implement a fixed compression ratio image compression method and an image decompression method.

The image compression method with a fixed compression ratio compresses the original data of the image into an image bit stream at a predetermined ratio. The image compression method with a fixed compression ratio includes: dividing the image into a plurality of image rows, wherein each image row includes at least one image row; specifying a storage space in the memory according to the compression ratio and the length of a bit stream of the original data; The compression ratio and the length of the bit stream of the original data divide the storage space into multiple compressed rows, which correspond to the image row respectively; record the starting position of each compressed row as a current end position; receive the pictures sequentially columns; compressing these image columns into a plurality of compressed bit streams respectively; A writing process sequentially writes the compressed bit stream into the corresponding compressed row.

The writing procedure performed by the encoder may include: receiving a compressed bit stream; writing the compressed bit stream into the corresponding compressed bit stream from the current end position of the corresponding compressed bit stream according to the field and image bank corresponding to the compressed bit stream area; and when the image row corresponding to the compressed bit stream is not the last image row of the image row to which it belongs, update the current end position of the corresponding compressed row area. The "last image sequence of the image row to which it belongs" refers to the last image sequence received according to the field order in the image row to which the image sequence belongs. The writing procedure may further include: when the image row corresponding to the compressed bit stream is the last image row of the associated image row, filling the corresponding compressed row area with a filling code after the compressed bit stream to be written.

The video decompression method is used to decode the video bitstream compressed by the encoder. The image decompression method includes: receiving at least one compressed row of image bitstreams; and decompressing the compressed bitstreams in the compressed row to a map sequence in the image row corresponding to the compressed row according to the field order.

In the image decompression method, the step of "decompressing these compressed bit streams in the compressed arrangement area back to the image sequence in the image arrangement corresponding to the compressed arrangement area" may include: calculating the respective corresponding bit streams in the compressed arrangement area A plurality of picture sequences of the fields; and decompressing the compressed bit streams in the compressed arrangement area to the image arrangement corresponding to the compressed arrangement area according to the order of the fields and the number of the image arrangements.

According to an embodiment of the present invention, the electronic device may further include at least a memory bus and an application module. The memory bus is electrically connected between the codec envelope module and the memory. The application module is electrically connected to the memory bus, and accesses at least one image row decompressed by the decoder through the memory bus. In addition, the electronic device may also include an on-chip random access memory (OCRAM) to record the current end positions of the compressed rows.

An image compression method with a fixed compression ratio is used to compress an original data of an image generated by a multi-field sensor into an image bit stream at a compression ratio. The image includes a plurality of image columns, and the multi-field sensor The detector outputs these image sequences in a field sequence, and each image sequence corresponds to a field index. The image compression method with a fixed compression ratio includes:

dividing the image into a plurality of image rows, wherein each image row includes at least one of the image columns;

Designate a storage space in a memory according to the compression ratio and a bit stream length of the original data;

According to the compression ratio and the length of the bit stream of the original data, the storage space is divided into a plurality of compressed row areas, and the compressed row areas correspond to the image rows respectively;

record the start position of each compressed area as a current end position;

receiving the graphs sequentially;

Compressing the graphs into a plurality of compressed bitstreams respectively; and

According to the group consisting of the field indexes of the image columns, the field order output by the multi-field sensor, and the image rows corresponding to the image columns, a writing process is used to sequentially write The compressed bit streams are written into the corresponding compressed rows.

The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments, but not as a limitation of the present invention.

Description of drawings

FIG. 1 is a schematic diagram of an electronic device according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of an image row according to an embodiment of the present invention;

3 is a flowchart of an image compression method with a fixed compression ratio according to an embodiment of the present invention;

FIG. 4 is a flowchart of a writing program according to an embodiment of the present invention;

5A-5F are memory schematic diagrams of a writing process according to an embodiment of the present invention;

FIG. 6 is a flowchart of step S106 according to an implementation example of the present invention;

FIG. 7A is a flow chart of image decompression according to an embodiment of the present invention;

FIG. 7B is a flow chart of image decompression according to another embodiment of the present invention; and

FIG. 7C is a flow chart of image decompression according to yet another embodiment of the present invention.

Among them, reference signs

20 electronic devices

22 Multi-field sensor

24 codec wrapper modules

241 encoder

242 decoder

25 memory

26 memory bus

27 On-chip random access memory

28 application modules

30 images

32, 32a, 32b image row

34, 34_1 ~ 34_m picture series

40 storage space

42a, 42b compression row area

44a, 44b current termination position

46 compressed bitstream

48 padding code

Detailed ways

Below in conjunction with accompanying drawing, structural principle and working principle of the present invention are specifically described:

The present invention provides an image compression method with a fixed compression ratio, an image decompression method and an electronic device thereof, so as to convert a raw data (raw data) of an image generated by a multi-field sensor Compress into an image bit stream with a compression ratio, or decompress (decode) the image bit stream back to the original data.

First, please refer to FIG. 1 , which is a schematic diagram of an electronic device according to an embodiment of the present invention. The electronic device 20 includes a multi-field sensor 22 , a codec wrapper module 24 and a memory 25 . The codec wrapper module 24 may include at least one encoder 241 for performing an image compression method with a fixed compression ratio, or/and at least one decoder 242 for performing an image decompression method. The electronic device 20 can be a digital camera, a mobile phone with a camera function, or a notebook computer with a camera function, etc., which can capture images with a multi-field sensor 22 and have an encoder 241 or/and a decoder 242 installation.

The multi-field sensor 22 can be, for example, a charge coupled device (CCD), which generates raw image data in a field order. Please also refer to FIG. 2 , which is a schematic diagram of an image row according to an embodiment of the present invention. The image 30 includes a plurality of lines 34, and the multi-field sensor 22 outputs all the lines 34 in field order. Each image row 34 corresponds to a field index (fieldindex), representing which field the image row 34 belongs to. Taking the implementation example of FIG. 2 as an example, the multi-field sensor 22 divides the image 30 into three fields, and outputs the image sequence 34 according to the field order of field 1 , field 2 , and field 3 . Therefore, the multi-field sensor 22 divides the image sequence 34 into 34_1, 34_4...34_n-2, 34_n+1...34_m-2, 34_2, 34_5...34_n-1, 34_n+2...34_m-1, 34_3 according to the field order. , 34_6...34_n, 34_n+3...34_m are output sequentially, where n and m are positive integers that are multiples of 3, and m is greater than n.

However, this description is only taken as an example, and does not limit the number of fields divided by the multi-field sensor 22 , the number of a column, and the order of fields output. The number of image columns 34 corresponding to each image field in an image band 32 is referred to as the number of image columns.

The memory 25 can be divided into different areas to separately store the compressed image bitstream and the result of decompressing the image bitstream. The electronic device 20 can also have at least one memory bus 26 electrically connecting the codec wrapper module 24 with the memory 25 . The memory 25 can be, for example, a double data rate memory (Double Data Rate memory, DDR memory), a dynamic random access memory (Dynamic Random Access Memory, DRAM) or a flash memory (Flash memory).

Please refer to FIG. 3 , which is a flowchart of an image compression method with a fixed compression ratio according to an embodiment of the present invention. First, the image 30 is divided into a plurality of image rows 32, wherein each image row 32 includes at least one image row 34 (step S101). For example, in FIG. 2 , the image 30 is divided into two image rows 32a and 32b; and the image rows 34_1˜34_n and the image rows 34_n+1˜34_m belong to the image rows 32a and 32b respectively. However, the image 30 can also be divided into more image rows 32 as required.

The encoder 241 allocates a storage space in the memory 25 according to the compression factor and the length of a bit stream of the original data (step S102 ). According to the compression ratio and the length of the bit stream of the original data, the storage space is divided into a plurality of compressed row areas, and these compressed row areas correspond to the image row 32 respectively (step S103 ). And the encoder 241 records the starting position of each compressed extent as a current ending position (step S104).

The electronic device 20 may further include an on-chip random access memory (OCRAM) 27 for storing the current end position of each compressed row. The on-chip random access memory 27 can be used by other modules inside the electronic device 20 in addition to storing the current end position. In addition, the electronic device 20 may also configure a random access memory or a plurality of registers in the codec wrapper module 24 for storing the current termination position.

Then the encoder 241 sequentially receives all the pictures 34 from the multi-field sensor 22 (step S105 ), and compresses these pictures 34 into a plurality of compressed bit streams (step S106 ). And according to the field index of each picture row 34, the field sequence output by the multi-field sensor 22 and the group formed by the image row 32 corresponding to the picture row 34, the encoder 241 uses a writing procedure to sequentially Write the compressed bit stream into the corresponding compressed row (step S107). In more detail, the codec wrapper module 24 is to repeatedly receive a picture sequence 34, compress the picture sequence 34 into a compressed bit stream, and write the compressed bit stream into the compressed row area with a writing program until the entire original The data is compressed and written into the storage space to become the image bitstream. The image bit stream is the data that fills the storage space of the compressed bit stream and the padding code.

Please refer to FIG. 4 , which is a flowchart of a writing procedure according to an embodiment of the present invention. The writing program first receives a compressed bit stream (step S111), and according to the field corresponding to the compressed bit stream and the image row 32, the compressed bit stream is compressed from the current end position of the compressed row area corresponding to the image row 32. The bit stream is written into the corresponding compressed row (step S112). In other words, the writing program determines the field index and the image row 32 to which the image sequence 34 compressed into the compressed bit stream belongs.

According to an exemplary implementation of the present invention, step S112 may include: calculating the fields corresponding to the compressed bit stream (namely field index) according to the order of fields and the number of picture columns 34 corresponding to each field; The number of picture rows 34 included in one video row 32 and the number of picture rows 34 corresponding to each field are used to calculate the video row 32 corresponding to the compressed bit stream.

After the compressed bit stream is written, the writing program judges whether the picture sequence 34 corresponding to the compressed bit stream is the last picture sequence 34 of the corresponding image row 32 (step S113 ). The "last image row 34 of the image row 32 to which it belongs" refers to the last image row 34 scanned and received in the image row 32 to which the image row 34 belongs according to the field order. When the picture row 34 corresponding to the compressed bit stream is not the last picture row 34 of the corresponding image row 32, update the current end position of the corresponding compressed row (step S114). And when the picture row 34 corresponding to the compressed bitstream is the last picture row 34 of the associated image row 32, a padding code is used to fill the corresponding compressed row area after the written compressed bitstream (step S115).

In step S115 , the codec wrapper module 24 determines whether there is still room in the compressed row after the last compressed bitstream is written in the compressed row. If so, the codec wrapper module 24 adds padding codes. Therefore, the starting position of each compressed row area is to store the compressed bit stream of the first image sequence 34 scanned in the field order in an image row 32 .

The image content in Figure 34 will affect the compressed result, so the length of the padding code in each compressed row may be different. In the stuffing code, all bits may have the same value, for example, all 0s or all 1s. The values in the pad code can also be different. The padding code may also be a verification code of the additional compressed bit stream, for example, it may be a check sum (check sum) of the corresponding compressed bit stream.

Please refer to FIG. 5A to FIG. 5F in conjunction with FIG. 4 , which are schematic diagrams of a memory writing process according to an embodiment of the present invention.

As shown in FIG. 5A, after step S104, the storage space 40 in the memory 25 is divided into compressed row areas 42a and 42b (hereinafter collectively referred to as compressed row areas 42) according to the image rows 32a and 32b. And the start positions of the compressed row areas 42a and 42b are respectively recorded as the corresponding current end positions 44a and 44b (hereinafter collectively referred to as the current end positions 44 ). Wherein, the starting positions of the compressed row areas 42 a and 42 b can be calculated from the sizes of the compressed row areas 42 a and 42 b and the starting position of the storage space 40 .

Since the field order is field 1, field 2, and field 3, the first compressed bitstream 46 received by the writing program corresponds to the first image sequence 34_1 of the image 30, and belongs to the first image sequence 34_1. 32a and map field 1. The write program writes the compressed bitstream 46 into the compressed row 42a corresponding to the image row 32a, and updates the current end position 44a of the compressed row 42a to be the next bit (bit) of the compressed bitstream 46 written. position, as shown in Figure 5B.

The encoder 241 sequentially writes the compressed bit streams 46 corresponding to the columns 34_1 , 34_4 . When the writing program receives the compressed bit stream 46 corresponding to the image row 34_n+1, it can be known in step S112 that it belongs to the image row 32b. Therefore, the compressed bit stream 46 corresponding to the column 34_n+1 is written into the compressed row area 42b according to the current end position 44b, as shown in FIG. 5C.

It is worth noting that the encoder 241 sequentially compresses all the pictures 34 belonging to the field 1 in the image row 32a and writes them into the compressed row 42a, and then compresses all the pictures 34 belonging to the field 1 in the image row 32b 34 are sequentially compressed and written into the compressed row 42b, and then sequentially written into the corresponding compressed row 42 according to the field order.

According to the above steps, after the image row 34 with the field index 3 in the image row 32 a is processed, the codec wrapper module 24 writes the padding code 48 . The encoder 241 then processes the image row 34 corresponding to the field index 3 in the image row 32 b, as shown in FIG. 5D , FIG. 5E and FIG. 5F .

Please refer to FIG. 6 , which is a flowchart of step S106 according to different implementation examples of the present invention. According to an embodiment of the present invention, step S106 may include: dividing the image sequence 34 with a plurality of compression windows (step S1061); sequentially compressing each compression window into a sub-bit stream (step S1062); These sub-bitstreams of 34 are used as compressed bitstream 46 (step S1063).

Please refer to FIG. 7A , which is a flow chart of image decompression according to an embodiment of the present invention. The decoder 242 can receive at least one compressed row 42 of the image bitstream from the memory 25 (step S121); and then decompress all the compressed bitstreams 46 in the compressed row 42 back to the compressed row 42 according to the field order All the image columns 34 of the corresponding image row 32 (step S122).

The electronic device 20 can further include at least one application module 28 . The application module 28 can obtain the image row 32 decompressed by the decoder 242 through the memory bus 26, and perform display or various digital image processing (DIP) on it. Since the decoder 242 can provide a single video row 32, the application module 28 does not need to wait for the entire video bitstream to be decoded to start operation.

According to an exemplary embodiment of the present invention, step S122 may include: calculating the number of columns corresponding to fields in the compressed arrangement area 42; All the image rows 34 in the image row 32 corresponding to the compressed row 42 are decompressed back.

According to another embodiment of the present invention, when each picture sequence 34 is compressed with multiple compression windows, step S122 may include: decompressing the sub-bitstream in the compressed bitstream 46 according to the field order and the number of picture sequences Compress back to the corresponding compression window as the image row 34 in the image row 32 corresponding to the compressed row area 42 .

Please refer to FIG. 7B and FIG. 7C , which are flowcharts of image decompression according to different implementation examples of the present invention. If the decoder 242 finds the stuffing code 48 after decompressing the compressed bit stream 46, the stuffing code 48 in the compressed row area 42 can be directly deleted (step S124). If the padding code 48 is a verification code corresponding to the compressed queue area 42, the decoder 242 can first verify whether the compressed bit stream 46 in the compressed queued area 42 is If it is correct (step S123), then delete the filling code 48.

To sum up, the encoder compresses the original data of the image with a compression ratio, and can ensure that the compressed image bit stream saves enough storage space and memory access bandwidth. By reserving the compression row and adding padding codes, the encoder can write all the image sequences corresponding to the same image row into the same compression row regardless of the field order. When it is only necessary to decompress any image bank in the entire image, it is not necessary to decompress the entire image bit stream, but only need to decode the compressed bank area corresponding to the required image bank. Therefore, the image compression method with a fixed compression ratio solves the problem in the prior art that it is difficult to take into account the order of fields and the boundaries of image rows.

Certainly, the present invention also can have other multiple embodiments, without departing from the spirit and essence of the present invention, those skilled in the art can make various corresponding changes and deformations according to the present invention, but these corresponding Changes and deformations should belong to the scope of protection of the appended claims of the present invention.

Claims (11)

1. An image compression method with a fixed compression ratio, which is used to compress an original data of an image generated by a multi-field sensor into an image bit stream with a compression ratio, the image includes a plurality of image columns, the The multi-field sensor outputs these image sequences in a field order, and each image sequence corresponds to a field index. It is characterized in that the image compression method with a fixed compression ratio includes: dividing the image into a plurality of image rows, wherein each image row includes at least one of the image columns; Designate a storage space in a memory according to the compression ratio and a bit stream length of the original data; According to the compression ratio and the length of the bit stream of the original data, the storage space is divided into a plurality of compressed row areas, and the compressed row areas correspond to the image rows respectively; record the start position of each compressed area as a current end position; receiving the graphs sequentially; Compressing the graphs into a plurality of compressed bitstreams respectively; and According to the group consisting of the field indexes of the image columns, the field order output by the multi-field sensor, and the image rows corresponding to the image columns, a writing process is used to sequentially write The compressed bit streams are written into the corresponding compressed rows. 2. The image compression method with a fixed compression ratio according to claim 1, wherein the writing program comprises: receiving a compressed bitstream; According to the field and the image bank corresponding to the compressed bit stream, write the compressed bit stream into the corresponding compressed bank area from the current end position of the corresponding compressed bank area; and When the picture row corresponding to the compressed bit stream is not the last picture row of the image row to which it belongs, update the current end position of the corresponding compressed row area. 3. the image compression method of fixed compression ratio according to claim 2, is characterized in that, this writing program also comprises: When the picture row corresponding to the compressed bitstream is the last picture row of the image row to which it belongs, a filling code is used to fill the corresponding compressed row area after the compressed bitstream is written. 4. An image decompression method, suitable for decoding the image bit stream compressed into by the image compression method with a fixed compression ratio according to claim 1, characterized in that the image decompression method comprises: receiving at least one compressed row of the video bitstream; and According to the field order, the compressed bit streams in the compressed row are decompressed back to the pictures in the image row corresponding to the compressed row. 5. The image decompression method according to claim 4, wherein according to the field order, the compressed bit streams in the compressed row area are decompressed back to the image corresponding to the compressed row area The steps for the columns in the row include: Calculating the numbers of a plurality of map columns respectively corresponding to the fields in the compressed row area; and According to the field order and the numbers of the image rows, the compressed bit streams in the compressed row are decompressed back to the tiles in the image row corresponding to the compressed row. 6. An electronic device, characterized in that it comprises: A multi-field sensor generates a raw data of an image in a field order, wherein the image includes a plurality of image columns, each of which corresponds to a field index; a memory; and A codec envelope module, the codec envelope module includes: At least one encoder is used to compress the original data of the image into an image bit stream with a compression ratio, and the encoder performs the following steps: dividing the image into a plurality of image rows, wherein each image row includes at least one of the image columns; Designate a storage space in the memory according to the compression ratio and a bit stream length of the original data; According to the compression ratio and the length of the bit stream of the original data, the storage space is divided into a plurality of compressed row areas, and the compressed row areas correspond to the image rows respectively; record the start position of each compressed area as a current end position; receiving the graphs sequentially; Compressing the graphs into a plurality of compressed bitstreams respectively; and According to the group consisting of the field indexes of the image columns, the field order output by the multi-field sensor, and the image rows corresponding to the image columns, a writing process is used to sequentially write writing the compressed bitstreams into the corresponding compressed rows; and At least one decoder is used to decode the image bit stream, and the decoder performs the following steps: receiving at least one compressed row of the video bitstream; and According to the field order, the compressed bit streams in the compressed row are decompressed back to the pictures in the image row corresponding to the compressed row. 7. The electronic device according to claim 6, further comprising: at least one memory bus electrically connected between the codec wrapper module and the memory; and At least one application module is electrically connected to the memory bus, and accesses at least one image bank decompressed by the decoder through the memory bus. 8. The electronic device according to claim 6, further comprising: An on-chip random access memory records the current end positions of the compressed rows. 9. The electronic device according to claim 6, wherein the writing program executed by the encoder comprises: receiving a compressed bitstream; writing the compressed bitstream into the corresponding compressed row area starting from the current end position of the corresponding compressed bitstream according to the field and the image row corresponding to the image sequence compressed into the compressed bitstream; and When the picture row corresponding to the compressed bit stream is not the last picture row of the image row to which it belongs, update the current end position of the corresponding compressed row area. 10. The electronic device according to claim 9, wherein the writing program further comprises: When the picture row corresponding to the compressed bitstream is the last picture row of the image row to which it belongs, a filling code is used to fill the corresponding compressed row area after the compressed bitstream is written. 11. The electronic device according to claim 6, wherein the decoder performs the decompression of the compressed bit streams in the compressed row area back to the compressed bit streams corresponding to the compressed row area according to the field order. The steps of the columns in the image row include: Calculating the numbers of a plurality of map columns respectively corresponding to the fields in the compressed row area; and According to the field order and the numbers of the image rows, the compressed bit streams in the compressed row are decompressed back to the tiles in the image row corresponding to the compressed row.