CN1148067C - Data compressing method for complex image of synthetic apertre radar - Google Patents
Data compressing method for complex image of synthetic apertre radarInfo
- Publication number
- CN1148067C CN1148067C CNB01118552XA CN01118552A CN1148067C CN 1148067 C CN1148067 C CN 1148067C CN B01118552X A CNB01118552X A CN B01118552XA CN 01118552 A CN01118552 A CN 01118552A CN 1148067 C CN1148067 C CN 1148067C
- Authority
- CN
- China
- Prior art keywords
- data
- image
- sub
- masking
- complex image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Landscapes
- Image Processing (AREA)
Abstract
The present invention provides a compression method for plural image data of a synthetic aperture radar (SAR). Plural images of the SAR are respectively compressed by using a mould value of the SAR and a real part (imaginary part), so the phase information of the images can be reserved. Subblocks of the images are classified in a self-adaptive way on the basis of contents, the contents of the images are divided by adopting a covering plate, and the images are compressed according to the covering plate on the basis of an improved JPEG method of different parameters.
Description
Technical field
The present invention relates to a kind of method of complex image data compression, particularly a kind of method that is used for the compression of synthetic aperture radar complex image data.
Background technology
Synthetic aperture radar (Synthetic Aperture Radar is called for short SAR) is a kind of earth observation systems of utilizing microwave imaging, is a kind of high-resolution imaging radar system.Synthetic aperture radar is widely used in numerous areas such as disaster monitoring, resource exploration, ocean research, military surveillance.The initial data of synthetic aperture radar is implemented synthetic aperture through signal processing, realize the orientation to high-resolution, utilize the frequency modulation on pulse compress technique realize distance to high-resolution.
As shown in Figure 1, it is the block diagram of the signal processing of typical synthetic aperture radar.Comprise among this signal processing flow figure:
Signal S (t) 1 that radar receives is separated into I phase and Q phase signals by quadrature, be converted into digital signal through A/D converter 3 then, the I complex signal that digital signal addition mutually obtains with Q is mutually handled the complex image data 5 (the mould value of complex data is the energy of ground return, and the main value at the multiple angle of plural number is the phase place of ground return) that obtains synthetic aperture radar via matched filter 4.Subsequently, the complex image data of SAR is stored or is transmitted.
For synthetic aperture radar, the data transfer rate of its initial data equals the product of the quantizing bit number of pulse repetition frequency, echo-signal sample frequency and each sampled point.Thus, can know that the data transfer rate of synthetic aperture radar is quite high: the ERS-1 of emission in 91 years and the satellite-borne SAR data transfer rate on the ERS-2 satellite can reach 105Mb/s, and the original data volume of every day reaches nearly 200GB.Huge data volume all constitutes suitable burden to the storage and the transmission of radar image data.The storage of data and transmission become the bottleneck of SAR system, have influence on the performance of whole system.Therefore,, under the prerequisite of the quality that guarantees the SAR image, reduce data transfer rate, have great importance for polarization sensitive synthetic aperture radar system by data compression technique.
Summary of the invention
The object of the present invention is to provide a kind of compression method of SAR complex image data, make SAR view data rate and data volume descend greatly, the SAR view data after guarantee rebuilding simultaneously still can satisfy and comprises that interference SAR handles the application requirements that obtains three-dimensional ground image.
Another object of the present invention provides a kind of method of complex image (hologram image) data compression, for the transmission of hologram image provides technical foundation.
To achieve the above object, the method for a kind of complex image compression, comprising two parts of compression and decompression:
1) data compression of complex image comprises following steps:
A. the mould value of calculated complex data quantizes with quantization table then, encodes for the data after quantizing, and obtains packed data;
B. get the real part or the imaginary part of complex data, quantize with quantization table then, encode for the data after quantizing, obtain packed data, two kinds of coded datas are stored with the corresponding quantization table or transmit;
2) decoding of packed data also is divided into two parts:
C. the mould Value Data of having encoded is decoded, decoding back data are divided into two-way, and one the tunnel directly exports, and is common view data; Another road calculates original mould Value Data according to the corresponding quantization table;
D. the real part or the imaginary data of having encoded are decoded, the decoding back calculates original real part or imaginary data according to the corresponding quantization table, then together with the original mould value information that recovers, calculates the phase information that the original complex image comprises.
The core of image compression is a coding module.Orthogonal transform coding can effectively be removed the correlation of picture signal, thus under the identical precision of images compressing image data significantly, such as existing Joint Photographic Experts Group.
Description of drawings
Specifically describe method of the present invention with reference to the accompanying drawings, wherein:
Fig. 1 is a synthetic aperture radar signal processing block diagram;
Fig. 2 is a complex image compression/de-compression theory diagram;
Fig. 3 is the Compression Coding Principle block diagram;
Fig. 4 is the decompression theory diagram;
Fig. 5 A is the mould value image of complex image data;
Fig. 5 B is the real part image of complex image data;
Fig. 5 C is the imaginary part image of complex image data;
Fig. 6 is that precision is 8 * 8 masking-out;
The mould value image of Fig. 7 for rebuilding;
The real part image of Fig. 8 for rebuilding;
The phase information of Fig. 9 for rebuilding.
Embodiment
As shown in Figure 2, for taking following compression process through the SAR complex image data that obtains after the signal processing:
The data compression 20 of complex image 21 is divided into two parts:
A. the mould value 23 of calculated complex data quantizes 25 then, encodes 27 for the data after quantizing, and obtains packed data.Coded data is stored with quantization table 26 or transmits.
B. get the real part (imaginary part) 22 of complex data, quantize 24 then, encode 28, obtain packed data for the data after quantizing.Coded data is stored with quantization table 29 or transmits.
The decompression 30 of packed data also is divided into two parts:
A. the mould Value Data of having encoded is decoded 32, decoding back data are divided into two-way, and one the tunnel directly exports, and is common view data; Another road calculates original mould Value Data 31 according to the corresponding quantization table, together with original real part (imaginary part) information of recovering 34, draws the phase information that the original complex image comprises by calculating its antitrigonometric function then.
B. real part (imaginary part) data 28 of having encoded are decoded 33, decoding calculates original real part (imaginary part) data 34 according to corresponding quantization table 29, together with the original mould value information that recovers, draw the phase information that the original complex image comprises then by calculating its antitrigonometric function.
Coded portion is improved on the basis of existing Joint Photographic Experts Group, the theory diagram of coded portion as shown in Figure 3:
Gray level image 51 is divided into sub-piece 52, carry out analysis and evaluation 58 for the information that each sub-image comprises, then with the sub-piece of image according to the result of assessment according to the significance level of information (for the SAR image, the information that large stretch of lower zone of gray value comprises is inessential relatively) classify, draw image classification information 59.Work out corresponding quantization table 510 according to the information of image segmentation, thereby realize the different and different content-based self-adapting compressing of the information that comprises according to image.
The process of whole compression is equivalent to piece image is divided into the mutual nonoverlapping subimage of several width of cloth according to the information that comprises, and compresses respectively, at the end that decompresses each number of sub images is carried out decompress(ion) then, and restoration and reconstruction original image then superposes.
The concrete steps of compression are as follows: 1. pair gray level image carries out the division 52 of 8 * 8 sub-pieces (size of sub-piece can change according to the needs of practical application).
1. the data that each sub-piece is comprised are analyzed, and calculate the average and the variance of its gray value, and the data that obtain according to analysis draw the classified information of this sub-piece.Specific practice is as follows: for the SAR image, calculate the average gray and the variance of sub-piece, for average gray in a certain gray value interval (such as 0～15) and its variance simultaneously also less than the class of returning into of variance threshold value (such as 20), image to this part adopts identical compression parameters compression, the classified information (corresponding to the corresponding jpeg algorithm compression quality factor) of noting this sub-piece with and coordinate position in entire image.(in diameter radar image, the zone that gray value is low is the shade of the water surface or massif, and the amount of information that wherein comprises is less relatively; And the thresholding of variance is used for guaranteeing not having tangible little target in this sub-piece because adopt the compression parameters of high compression ratio loss of information.)
2. make corresponding masking-out according to the coordinate of the sub-piece of the same classification that obtains,, constitute this classification subimage with view data.Masking-out be one with big bitmap images (black and white image, value are 0 and 1) such as original image, corresponding each classification, the pixel value at place, position that belongs to the sub-piece of this class is 1, other regional value is 0.
3. the masking-out of original image and each classification is carried out the operation of logical, draw the subimage of each classification.
For each the classification subimage according to the given compression parameters of its classified information, the improved jpeg algorithm of antithetical phrase image applications is encoded:
A. discrete cosine transform 54.
B. calculate (or searching) corresponding quantization table 510 according to classified information.The key component of jpeg algorithm is the selecting for use of quantization table in the algorithm: quantization step is big, and compression ratio is just high, and picture quality is relatively just poor; Quantization step is little, and compression ratio is just low, and picture quality is all right relatively.The selecting for use of quantization table for each classification subimage: can specify a fixing quantization table, then different classification be multiply by different coefficient calculations with this quantization table and obtain; Also can carry out statistical analysis to similar view data in advance, calculate and put out the quantization table of optimizing targetedly that is fit to employing for the classification subimage of different characteristic in order.
C. quantize 55.
D. entropy coding 56, preserve corresponding Huffman table.
5. the coding 57 after will compressing is preserved or transmission together with sub-block sort information, quantization table and Huffman table.
The decode procedure of compressing image data is opposite with cataloged procedure, its theory diagram as shown in Figure 4:
The concrete steps that decompress are as follows:
1. from packed data 61, obtain sub-block sort information 67.
2. according to sub-block sort information, from packed data, obtain corresponding Huffman table 68 and quantization table 69 data.
For each the classification subimage according to relevant parameters, the jpeg algorithm of application enhancements is decoded:
A. carry out the entropy decoding according to corresponding Huffman table.
B. go to quantize 63 according to corresponding quantization table.
C. inverse discrete cosine transform 65.
4. according to the masking-out information of the classified index correspondence that obtains,, rebuild this classification subimage 66 with decoded view data.
5. the superpose subimage of each classification, reconstructed image information.
Illustrate:
The phase information that comprises in the synthetic aperture radar plural number information is mainly used in generating the digital elevation image of ground form, and we are that the SAR complex image data of 512 * 512 synthetic aperture radar massif is an example with size.Show the real part of complex image and absolute value images such as Fig. 5 A of imaginary part respectively.Real part and imaginary part (positive and negative real number) with identical ratio, are transformed to the codomain (0～255) of gray level image, as Fig. 5 B and 5C.
To this plural number, we set the branch block size is 8 * 8, makes the masking-out of 8 * unit 8 size.According to image information, we are divided into two classes with sub-piece, draw masking-out image such as Fig. 6.
According to masking-out, select different compression quantization tables, compression molding value image.Decompress(ion) contracts according to the image of masking-out comprehensive redevelopment as shown in Figure 7.
According to masking-out the real part view data is compressed equally, the image of restoration and reconstruction then, as shown in Figure 8.
Thus, can calculate the phase information that reconstructed image comprises, as shown in Figure 9.
Compare with present existing method for compressing image, new method has increased the consideration of several aspects:
1. increased the consideration to the compression method of complex image, promptly not merely considered the compression as the plane picture of eye-observation, the phase information that comprises in the complex image is preserved in compression simultaneously.And phase information is for SAR, and especially interference SAR is a quite important or even indispensable part.This method also can be considered to be applied in the compression of the hologram image with similar requirement simultaneously.
2. for the reservation of phase information, we adopt the amplitude and real part (imaginary part) data of preserving each pixel of complex image, compress respectively.When decompressing, obtain the phase place of each pixel according to corresponding trigonometric function relation then.
3. for the compression of amplitude or real part (imaginary part) data, we quantize data becomes 0～255 integer, i.e. the gray level image of 256 gray scales, and the jpeg algorithm after the application enhancements compresses this gray level image then.
4. in the compression of gray level image, we at first carry out cutting apart of onesize sub-piece with gray level image, the information that the sub-image after cutting apart comprises are analyzed, according to the difference of analysis result taxonomic revision in addition, make masking-out, and provide corresponding D CT quantization table.
5. when gray level image being compressed, the object of compression is the subimage that image and masking-out are carried out the corresponding classification after logical is operated, and quantizes according to the quantization table of optimizing.In jpeg algorithm, compression ratio and quality are determined by the DCT quantization table, so just can adopt different quantization tables according to the difference of picture material: the DCT quantization table that the subimage that comprises important information is adopted thin quantization step; And the subimage that comprises inessential information adopts the DCT quantization table of thick quantization step, thereby realizes content-based self-adapting compressing.
Claims (11)
1, the method for a kind of complex image compression, comprising two parts of compression and decompression:
1) data compression of complex image comprises following steps:
A. the mould value of calculated complex data quantizes with quantization table then, encodes for the data after quantizing, and obtains packed data;
B. get the real part or the imaginary part of complex data, quantize with quantization table then, encode for the data after quantizing, obtain packed data, two kinds of coded datas are stored with the corresponding quantization table or transmit;
2) decoding of packed data also is divided into two parts:
C. the mould Value Data of having encoded is decoded, decoding back data are divided into two-way, and one the tunnel directly exports, and is common view data; Another road calculates original mould Value Data according to the corresponding quantization table;
D. the real part or the imaginary data of having encoded are decoded, the decoding back calculates original real part or imaginary data according to the corresponding quantization table, then together with the original mould value information that recovers, calculates the phase information that the original complex image comprises.
2. complex image compression method according to claim 1 is characterized in that described quantized data comprises step:
Complex image is divided according to the sub-piece of custom size;
The data that each sub-piece comprises are analyzed, calculated the average and the variance of gray value of the image of its mould, the data that obtain according to analysis draw the classified index of this sub-piece, and provide the DCT quantization table that this classification should be adopted.
3. complex image compression method according to claim 2 is characterized in that described sub-piece division comprises step:
Coordinate according to the sub-piece of classified index that obtains and same classification is made corresponding masking-out, with view data, constitutes this classification subimage;
4. complex image compression method according to claim 3 is characterized in that described each classification subimage according to the given compression parameters of its classified information, and antithetical phrase image applications compression algorithm is encoded to image.
5. method according to claim 3 is characterized in that the size of the sub-piece that the size of the minimum unit of described masking-out equals to divide image being combined with masking-out, according to the masking-out difference, subimage is carried out the compression and the decompression of image according to different parameters.
6. according to claim 3 or 5 described complex image compression methods, it is characterized in that each masking-out all corresponding in the different subimages classification, each masking-out is also divided according to the sub-piece of pre-sizing, in the masking-out of an image classification, the sub-piece of masking-out corresponding to the sub-piece of original image that belongs to this image classification is a logical one, otherwise the sub-piece of this masking-out is a logical zero.
7. complex image compression method according to claim 5, it is characterized in that by masking-out be compressed the subimage that image carries out obtaining after the logical operation corresponding classification, and quantize according to the quantization table that should classify mutually.
8. complex image compression method according to claim 3 is characterized in that adopting different quantization tables according to the classification of picture material is different: the quantization table that the subimage of the information that comprises application need is adopted thin quantization step; And the subimage that does not comprise the information of application need adopts the quantization table of thick quantization step.
9. complex image compression method according to claim 8 is characterized in that the importance of whether judging image subblock according to the average gray value and the variance of image subblock in preset range.
10. complex image compression method according to claim 1 is characterized in that described compression algorithm is the jpeg image compression method.
11. complex image compression method according to claim 1 is characterized in that described complex image is the synthetic aperture radar complex image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB01118552XA CN1148067C (en) | 2001-06-01 | 2001-06-01 | Data compressing method for complex image of synthetic apertre radar |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB01118552XA CN1148067C (en) | 2001-06-01 | 2001-06-01 | Data compressing method for complex image of synthetic apertre radar |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1390059A CN1390059A (en) | 2003-01-08 |
| CN1148067C true CN1148067C (en) | 2004-04-28 |
Family
ID=4663265
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB01118552XA Expired - Fee Related CN1148067C (en) | 2001-06-01 | 2001-06-01 | Data compressing method for complex image of synthetic apertre radar |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1148067C (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100386648C (en) * | 2005-02-01 | 2008-05-07 | 清华大学 | Real-time imaging method and real-time imager for synthetic aperture radar |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1307428C (en) * | 2004-04-09 | 2007-03-28 | 中国人民解放军国防科学技术大学 | Real-virtual relating method for generating interference phase pattern of synthetic aperture radar without interference spots |
| CN1982914B (en) * | 2005-12-15 | 2010-10-13 | 四川川大智胜软件股份有限公司 | Radar data self-adaptive non-loss compression based on decode and message discrimination |
| CN100464589C (en) * | 2006-02-07 | 2009-02-25 | 中国船舶重工集团公司第七○九研究所 | Full radar video compound technology |
| CN100541227C (en) * | 2006-03-31 | 2009-09-16 | 中国科学院空间科学与应用研究中心 | Subaperture radar altimeter |
| CN101135726B (en) * | 2007-09-21 | 2010-11-03 | 北京航空航天大学 | Satellite carried SAR inner marking signal processing platform system and realization method thereof |
| CN101655551B (en) * | 2008-08-20 | 2012-06-27 | 中国科学院电子学研究所 | Codec method and codec device for initial data compression of synthetic aperture radar with fixed frame size |
| CN101363911B (en) * | 2008-09-23 | 2011-05-11 | 清华大学 | Multi-viewer fully-polarized synthetic aperture radar data compression method |
| CN112766401B (en) * | 2021-01-28 | 2022-03-01 | 哈尔滨工业大学 | Countermeasure sample defense method based on significance countermeasure training |
| CN115914630B (en) * | 2023-01-06 | 2023-05-30 | 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) | Image compression method, device, equipment and storage medium |
| CN116106851B (en) * | 2023-04-04 | 2023-06-20 | 中国科学院空天信息创新研究院 | Method and device for compressing raw data of synthetic aperture radar |
-
2001
- 2001-06-01 CN CNB01118552XA patent/CN1148067C/en not_active Expired - Fee Related
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100386648C (en) * | 2005-02-01 | 2008-05-07 | 清华大学 | Real-time imaging method and real-time imager for synthetic aperture radar |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1390059A (en) | 2003-01-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP1591963B1 (en) | Adaptive quantisation of a depth map | |
| CN110225341A (en) | A kind of code flow structure image encoding method of task-driven | |
| CN101420614B (en) | Image compression method and device integrating hybrid coding and wordbook coding | |
| CN1148067C (en) | Data compressing method for complex image of synthetic apertre radar | |
| CN1906948A (en) | Image coding apparatus and method for predicting motion using rotation matching | |
| CN1183491C (en) | Method and system for compressing image data with wavelet transform and quadtree coding for telecommunication | |
| RU2567988C2 (en) | Encoder, method of encoding data, decoder, method of decoding data, system for transmitting data, method of transmitting data and programme product | |
| CN1212806A (en) | Progressive image coding | |
| Zanuttigh et al. | Compression of depth information for 3D rendering | |
| CN103179397A (en) | Hyperspectral image compressing and restructuring device and method | |
| CN101631243A (en) | Image encoding/decoding method based on wavelet transformation | |
| CN1284120C (en) | Synthetic aperture radar complex numeric image data real time automatic compression method | |
| CN1628466A (en) | Context-sensitive encoding and decoding of a video data stream | |
| Boopathiraja et al. | Significance of image compression and its upshots-A survey | |
| CN109672891B (en) | Lossless secondary compression method of JPEG image | |
| Shingate et al. | Still image compression using embedded zerotree wavelet encoding | |
| Sandhu et al. | Matlab Based Image Compression Using Various Algorithms | |
| CN1822050A (en) | Full phase anticosine double orthogonal transformation and its improving method for JPEG | |
| Dinh et al. | Side information generation using extra information in distributed video coding | |
| Pancholi et al. | Tutorial review on existing image compression techniques | |
| Mekhalfa et al. | A Lossless hybrid wavelet-fractal compression for welding radiographic images | |
| Kumar et al. | Comparative analysis of wavelet based compression methods | |
| Shu et al. | Arbitrarily shape object-based video coding technology by wavelet-based contourlet transform | |
| Thayammal et al. | Image compression using multidirectional anisotropic transform: Shearlet transform | |
| Pareek et al. | An Image Compression Using Multilayer Wavelet Transform with 2DTCWT: A Review |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| ASS | Succession or assignment of patent right |
Owner name: XIAOSHAN INDUSTRY INSTITUTE Free format text: FORMER OWNER: INSTITUTE OF COMPUTING TECHNOLOGY, CHINESE ACADEMY OF SCIENCES Effective date: 20130401 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 100080 HAIDIAN, BEIJING TO: 311215 HANGZHOU, ZHEJIANG PROVINCE |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20130401 Address after: 311215, Hangzhou, Zhejiang province Xiaoshan District, Ning Town, civilization Road, No. 328, Jiangning building, east block, 7 floor Patentee after: XIAOSHAN INDUSTRY RESEARCH INSTITUTE Address before: 100080 Haidian District, Zhongguancun Academy of Sciences, South Road, No. 6, No. Patentee before: Institute of Computing Technology, Chinese Academy of Sciences |
|
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20040428 Termination date: 20160601 |