CN103347185A - Unmanned aerial vehicle scouting image synthesis compressed encoding method based on selective block transformation - Google Patents
Unmanned aerial vehicle scouting image synthesis compressed encoding method based on selective block transformation Download PDFInfo
- Publication number
- CN103347185A CN103347185A CN2013102653703A CN201310265370A CN103347185A CN 103347185 A CN103347185 A CN 103347185A CN 2013102653703 A CN2013102653703 A CN 2013102653703A CN 201310265370 A CN201310265370 A CN 201310265370A CN 103347185 A CN103347185 A CN 103347185A
- Authority
- CN
- China
- Prior art keywords
- image
- coding
- compression
- aerial vehicle
- unmanned aerial
- 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.)
- Granted
Links
Images
Landscapes
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention discloses an unmanned aerial vehicle scouting image synthesis compressed encoding method based on selective block transformation. The method comprises the steps of synthesizing the H.264 compressing scheme and the JPEG 2000 compressing scheme, carrying out selective algorithm adjustment and improvement on block transformation in the H.264 video coding standard, and obtaining a compressed code stream by means of compressed encoding. According to the unmanned aerial vehicle scouting image synthesis compressed encoding method based on the selective block transformation, the problem of limitation on image transmission bandwidth caused by limitation of size, weight and power consumption and other respects of an unmanned aerial vehicle platform, and influence of channels, external environment and enemy interference and other factors on actual image information transmission is effectively solved, uncompressed image source information is differentiable under the situation that still images before and after compression are similar in respect of PSNR and subjective quality evaluation of a scouting video is high, the compression ratio is greatly increased, and the scheme is systematic.
Description
Technical field
The invention belongs to the digital video image processing technology field, be specifically related to a kind of comprehensive compaction coding method of unmanned plane reconnaissance image based on the conversion of selectivity piece.
Background technology
Unmanned plane portability multiple-task load, work under various modes, its scouting method has multiple, except visible image capturing is obtained video, also has that digital photograph is taken pictures, infrared imaging, multispectral scanner, synthetic aperture radar etc.Modern unmanned plane reconnaissance image all adopts the digitlization transmission, therefore the amount of information of reconnaissance image is very big, and want real-time Transmission, simultaneously, image information can be subjected to channel, external environment and enemy's interference and the influence of other factors in the actual transmissions process, owing to the restriction of unmanned plane platform at aspects such as volume, weight and power consumptions, the image transmission bandwidth is limited again, and therefore the suitable unmanned plane of a kind of high compression ratio of needs satisfies the universal compressed method of application demand under event of bandwidth limitation.
The view data that derives from the unmanned plane scouting trends towards multi-sourceization, though the method for image compression and algorithm are a lot, can reach high compression ratio and fidelity requirement in some applications.But the universal compressed algorithm that can be suitable for all kinds of unmanned plane reconnaissance images does not have.Be example with dynamic video and rest image, use the more and significant method of compression effectiveness at present and be based on H.264(dynamic) and the compression method of JPEG2000 standard (static state).Wherein, H.264 Bian Ma basic step is: the macro block that image is divided into a plurality of 16x16 pixel, motion component with each each frame of block motion vector representative image sequence, type (I frame according to frame, P frame or B frame) selection intraframe coding or interframe encoding mode, coded frame or reference frame carry out estimation and compensation generation forecast macro block to subsequent frame to use the front, predicted macroblock P and current macro are subtracted each other, obtain the image residual block, pass through integer DCT(discrete cosine transform then), quantize, reorder, cataloged procedure is finished in the processing of part such as entropy coding; The basic step of JPEG2000 coding is: earlier source image data is carried out the preliminary treatment in early stage, again the entire image after handling is carried out (9,7) DWT(wavelet transform), then to the wavelet coefficient after the conversion quantize, Bit-Plane Encoding, layering assembling and packing, form the output code flow of JPEG2000 standard at last.
At present, in the middle of the unmanned plane compressibility, adopt one of following two kinds of schemes usually: discrete scheme, dynamic video adopt H.264 compression scheme, and rest image then adopts the JPEG2000 compression scheme to compress, and can not use under airborne resource-constrained situation; Using H.264, algorithm compresses two types image, when compressing the rest image frame like this, relative DWT transform method, H.264 the integer dct transform that adopts in is applicable to the fritter conversion as the 4x4 size, though the mismatch problems of having avoided the general 8x8 discrete cosine transform used in the standard in the past and inverse transformation often to occur, well kept the details of necessity in the image, but at the piece conversion bigger as 16x16, adopt integer DCT not only to take more memory space, also piecemeal and ringing effect may occur.
Summary of the invention
The objective of the invention is in order to address the above problem, compressed encoding demand at the unmanned plane reconnaissance image, the comprehensive compaction coding method of unmanned plane reconnaissance image based on the conversion of selectivity piece has been proposed, comprehensively H.264 with the JPEG2000 compression standard, the piece transform method is carried out categorizing selection based on block size, thereby realize the comprehensive compressibility of two class images.
The comprehensive compaction coding method of unmanned plane reconnaissance image based on the conversion of selectivity piece comprises the following aspects:
(1) macroblock partitions in the scheme H.264;
Coded image is divided by standard H.264, be divided into the macro block of 16 * 16 pixel sizes.
Coded image comprises video information, SAR image, high spectrum image.
(2) according to type selecting intraframe coding or the interframe encoding mode of picture frame, carry out predictive coding respectively;
If the multiframe dynamic video adopts inter prediction and intraframe predictive coding; If the high spectrum image after the SAR image of single frames or the removal spectrum intersexuality then only adopts intraframe predictive coding, obtain the predicted value of current macro;
(3) the image residual block calculates;
To subtract each other through macroblock prediction value and the former macro block that step (2) obtain, obtain the image residual block;
(4) based on piece size Selection spatial transform method;
16 * 16 and 8 * 8 luminance pixel pieces are carried out (9,7) wavelet transform, and all the other macro blocks then carry out discrete cosine transform;
(5) entropy coding:
Residual error data, motion vector information after control information and the quantification are carried out data compression.
After finishing the entropy coding, obtain the compressed bit stream of compressed image, use for transmission and storage through the network self-adapting layer.
The invention has the advantages that:
(1) comprehensively H.264 with these two kinds of compression schemes of JPEG2000, carry out different transform methods at the different masses size, combine the advantage of DCT and DWT conversion;
(3) (9,7) DWT is replaced the H.264 dct transform of middle part piecemeal, eliminated the blocking effect when compressed image recovers, satisfying under the certain condition situation, improved compression ratio.
Description of drawings
Fig. 1 is method flow diagram of the present invention;
Fig. 2 is H.264 transcoding, coding transform and quantizing process flow chart;
Fig. 3 is of the present invention shift process figure;
Fig. 4 is (9,7) DWT boosting algorithm structure chart among the present invention.
Embodiment
Below in conjunction with accompanying drawing the specific embodiment of the present invention is elaborated.
Comprehensive existing compaction coding method and standard, in conjunction with H.264 with these two kinds of compression standards of JPEG2000, piece transform method is H.264 improved, foundation is compressed image based on the comprehensive compaction coding method of unmanned plane reconnaissance image of selectivity piece conversion.
A kind of comprehensive compaction coding method of unmanned plane reconnaissance image based on the conversion of selectivity piece of the present invention, flow process comprises following step as shown in Figure 1:
The first step, the macroblock partitions in the scheme H.264, flow process specifically comprises as shown in Figure 2:
H.264 the intra prediction mode of introducing in, considered the spatial redundancy of regular object and regular background in the single width video image, intra-frame prediction method based on two kinds of block sizes 4 * 4 and 16 * 16 has been proposed, wherein especially effective to flat site based on 16 * 16 infra-frame prediction, and because the unmanned plane during flying height is higher, speed is fast, the airborne sensor angle of visual field is bigger, so the interested object pixel of institute is few in the unmanned plane reconnaissance image, in-frame correlation is poor than normal image, and frame-to-frame correlation is that the time correlation is not strong yet.
Coded image is divided by standard H.264, be divided into the macro block of several 16 * 16 pixel sizes.
Coded image comprises video information, SAR image, high spectrum image.
The macro block of 16 * 16 pixel sizes is made up of the colour element piece of 16 * 16 luminance pixel pieces and additional 8 * 8.
8 * 8 colour element piece comprises one 8 * 8 Cb(saturation) block of pixels and one 's 8 * 8 Cr(tone) block of pixels.
(2) according to type selecting intraframe coding or the interframe encoding mode of picture frame, carry out predictive coding respectively;
According to standard compression method H.264, according to the type of present frame in the current input macro block, carry out predictive coding respectively: if the multiframe dynamic video adopts inter prediction and intraframe predictive coding; If SAR image or the high spectrum image of single frames adopt intraframe predictive coding, obtain the predicted value of current macro.Adopt the Forecasting Methodology of former H.264 standard compression method in the frame with inter prediction.
After the predictive coding, obtain the block of pixels that several differ in size.
(3) will subtract each other through macroblock prediction value and the former macro block that step (2) obtain, obtain the image residual block.
(4) the image residual block is carried out (9,7) DWT and integer dct transform and quantification.
Transition coding and quantification technique come the frequency domain correlation in the removal of images signal and reduce the dynamic range of image coding, thereby reach the purpose of further compression.As shown in Figure 2, be H.264 transcoding, coding transform and quantizing process flow chart, H.264 in macroblock size be 16 * 16, the piece of each 4 * 4 size is wherein carried out 4 * 4 integer dct transform after, obtain 16 4 * 4 transformation matrix.Integer DCT has avoided the encoding and decoding mismatch problem in the standard in the past, make inverse transformation unbalance problem can not occur, and its computing only comprise the plus-minus and the displacement, and will quantize the fusion therein, effectively reduce operand, improved the real-time of image compression.Though dct transform calculates simple, and Fast implementation is arranged, recovering image when compression has tangible blocking effect, is not suitable for carrying out relatively large conversion.
Wavelet transformation has the better energy centrality, the compression ratio height, compression speed is fast, can keep the feature of image constant substantially after the compression, can be anti-interference in transmittance process, and have the video localization property simultaneously, the feature of coefficient is conducive to coding after the conversion, so the coding method of wavelet transformation has better compression performance.Because it is less to carry out 4 * 4 prediction back residual values in H.264, be not suitable for carrying out wavelet transformation, and wavelet transformation is better at big macroblock coding transform effect, therefore the present invention is by utilizing H.264 video encoding standard, image 16 * 16 and 8 * 8 brightness block of pixels are adopted (9,7) DWT conversion, less residual block continue to use the integer dct transform, as shown in Figure 3.
Wavelet transformation has fabulous energy accumulating effect, but need make convolution when realizing, so complexity is higher.And the introducing of lifting model, the convolution operation when making wavelet transformation can be finished by lifting and stretching step, further reduces the computing of wavelet transformation, and it has following advantage:
1), this bit manipulation, saving internal memory;
2), utilize compound assignment, reduce the floating-point operation amount, speed is fast, efficient is high;
3), all operations in up step walks abreast, and is serial between a plurality of up step;
4), inverse transformation process is extremely simple;
Through a large amount of contrast experiment, to image realize the energy accumulating effect best be 9/7 wavelet basis, its low-frequency filter is 9 taps, high frequency filter then is 7 taps.The formula that its hoisting way is finished is as follows:
There are 4 to promote step (1-4) and two stretching steps (5-6) in the step of 9/7 Lifting Wavelet;
Y(2n+1)=X
est(2n+1)+(α×[X
ext(2n)+X
ext(2n+2)]) [Step1]
Y(2n)=X
ext(2n)+(β×[Y(2n-1)+Y(2n+1)]) [Step2]
Y(2n+1)=Y(2n+1)+(γ×[Y(2n)+Y(2n+2)]) [Step3]
Y(2n)=Y(2n)+(δ×[Y(2n-1)+Y(2n+1)]) [Step4]
Y(2n+1)=-K×Y(2n+1) [Step5]
Y(2n)=(1/K)×Y(2n) [Step6]
Wherein: x (n) is input signal, X
Ext(n) be the subsignal that the even sequence number sampling of x (n) is formed, α, beta, gamma, δ, k are constant.Each parameter is
And k=1.230174105 is inversely transformed into:
X(2n)=KxY
ext(2n)
X(2n+1)=-(1/K)×Y
ext(2n+1)
X(2n)=X(2n)-(δ×[X(2n-1)+X(2n+1)])
X(2n+1)=X(2n+1)-(γ×[X(2n)+X(2n+2)])
X(2n)=X(2n)-(β×[X(2n-1)+X(2n+1)])
X(2n+1)=X(2n+1)-(α×[X(2n)+X(2n+2)])
9/7 wavelet arithmetic structure as shown in Figure 4.Formed by three steps: division, prediction and renewal.Xe(n) be the subsignal that even sequence number sampling is formed, Xo (n) is the subsignal that strange sequence number sampling is formed.Original image is through behind the wavelet transformation, the view data total amount is without any minimizing, just the energy of entire image is redistributed, make in the large data sets of wide range in a less zone, and the dynamic range of data is very little in very big zone, for image compression has been created condition, but wavelet transformation itself and the image compression that is unrealized.Image compression is to be realized by the quantification of its back, coding.
In the quantification treatment stage, H.264 standard is used the scalar quantization device, and its quantization step has 52 grades, during use by QP value index.Notice the every increase by 6 of QP value, quantization step Qstep just doubles; Qstep increases progressively with 0.125.The expansion of range of quantization step makes encoder to control more flexibly and accurately, reaches compromise between bit rate and picture quality.
In encoder, the conversion coefficient of the 4x4 that each has quantized is mapped as the matrix of 16 elements with the Zig-Zag order, finishes and reorders.
(5) entropy coding
H.264/AVC the entropy coding that adopts in the standard is a kind of method of harmless compression, has utilized the statistical property of coded identification to carry out data compression.Two kinds of basic skills commonly used in the video coding are: variable-length encoding (Variable Length Coding, VLC) and arithmetic coding (Binary Arithmetic Coding, BAC).
Information such as the residual error data after H.264/AVC standard utilizes the entropy coding to control information and quantification, motion vector are carried out data compression.H.264/AVC the entropy coding in has two types.(Universal Variable Length Coding UVLC) comes all grammers stream elements except the scanning quantization parameter are encoded based on the consistent mutilation long codes of exponential type Golomb sign indicating number in employing; The scanning quantization parameter then adopt more flexibly based on contextual adaptive variable length coding (Content-based Adaptive Variable Length Coding, CAVLC), this also be H.264/AVC in the coded system of acquiescence.The correlation that CAVLC provides according to the symbol of having encoded is dynamically adjusted the code table of use in the coding and the renewal of hangover coefficient suffix length, has embodied adaptive characteristic, has obtained very high compression ratio.
H.264/AVC also provide a kind of based on contextual adaptive binary arithmetic coding (Content-based Adaptive Binary Arithmetic Coding, CABAC), as an option of main class (Main Profile).CABAC compares with the coded system CAVLC that H.264/AVC gives tacit consent to has higher code efficiency.Its advantage is that encoder does not need the priori of information source statistical property, but estimates adaptively in cataloged procedure.
After having finished the entropy coding, just obtained the compressed bit stream of compressed image, through NAL(network self-adapting layer) use for transmission and storage.
In second step, decompress.
Decode procedure is responsible for and will be met the H.264 compressed bit stream decoding of code stream standard, and carries out image reconstruction, is the inverse process of compression process.Decode procedure has utilized decoder H.264 from NAL(network self-adapting layer) receive the bit stream of compression, obtain a series of quantization parameter X through code stream being carried out the entropy decoding and reordering, these quantization parameters obtain residual error data D through inverse quantization and inverse transformation, wherein, the inverse transformation of quantization parameter is the inverse process of above-mentioned selectivity piece conversion, 16x16 and 8x8 luminance pixel piece use (9,7) the inverse transformation mode of DWT, all the other pieces then carry out idct transform, decoder uses the header that decoding obtains from code stream to create a fast PRED of prediction, PRED and residual error data D summation obtain image block data uf, last each image block data uf obtains the decoding block F of reconstructed image by deblocking filtering, finishes decompression process.
The present invention, at unmanned plane reconnaissance image characteristics, comprehensively H.264 with these two kinds of compression algorithms of JPEG2000, mainly be that 16x16 and 8x8 luminance block are adopted (9,7) DWT conversion, all the other macro blocks continue to use integer dct transforms.
Embodiment
Experiment is carried out at the VS2010 platform, at rest image, is guaranteeing under the close situation of PSNR, relatively its compression ratio size.At scouting video, because PSNR differs bigger, be chosen in subjective quality hardly under the affected situation, quantized value QP value is made as 20,51 respectively, and code check is 512kb/s, its compression ratio size of comparison.Single-frame images is selected high spectrum image PIC4.bmp, 1024x1024, big or small 3M; Dynamic video selects unmanned plane to scout video BGR_352x288.yuv, 352x288, and size is 18.9M.
1, image is divided into the 16x16 macroblock size;
2, at single-frame images PIC4, image is carried out infra-frame prediction; Carry out in the frame and inter prediction at scouting video BGR_352x288.yuv;
3, predicted value and observed value are subtracted each other to obtain the image residual error fast, and 16x16 in the image residual error and 8x8 luminance pixel piece are carried out (9,7) DWT transform and quantization, all the other pieces carry out dct transform and quantification;
5, the result in the previous step is carried out the entropy coding;
6, decompression obtains output file: generate the pic4.264 file respectively through the x264 program coding, and big or small 49.8KB and BGR_352x288.264 file, size is 129KB.
(wherein, the image quality evaluation standard that adopts in the following table is the SSIM(structural similarity), PSNR(Y-PSNR)) operation result is:
Wherein, rest image PIC4, last side data is scheme compression result H.264, and intermediate data is the JPEG2000 compression result, and following side data then is compression scheme test result of the present invention; In the dynamic video sequence BGR_352x288 test result, upside is scheme compression result H.264, and downside is compression scheme test result of the present invention, and test result is:
1) compare with current discrete scheme: compare with JPEG2000 compression rest image, under the situation of be more or less the same at PSNR (low 0.557dB), the compression ratio of this method has had significantly raising, about 284%; Compare with more H.264 compressing dynamic video, having higher rating, lost part PSNR(at subjective quality mainly is luminance component PSNR, low 2.987dB) situation under, the compression ratio of this method is greatly improved, about 25.6%;
2) compare with the current integrated scheme of taking, namely compress dynamically and static 2 kinds of images with standard H.264: the dynamic video aspect is as described in the epimere; Rest image is more or less the same at PSNR under the situation of (low 1.024dB), and the compression ratio of this method more H.264 compression scheme also is significantly improved, and about 14%.Therefore, this method can better meet the application demand of unmanned plane under event of bandwidth limitation.
Consideration based on the following aspects: H.264 in the compression standard, there is the possibility that can optimize lifting in the spatial alternation of large-size piece; The DWT conversion is applicable to the bulk conversion in the JPEG2000 compression standard, has concentration of energy, characteristics that compression ratio is high, the present invention proposes a kind of comprehensive compaction coding method of unmanned plane reconnaissance image based on the conversion of selectivity piece.This method synthesis H.264 with these two kinds of compression schemes of JPEG2000, in 16x16 and the conversion of 8x8 luminance block, result of the test showed during H.264 DWT be applied to:
1) compare with current discrete scheme: compare with JPEG2000 compression rest image, at the PSNR(Y-PSNR) under the situation of be more or less the same (about low 1dB), compression ratio be significantly improved (statistics on average can improve 10% after tested); Compare with H.264 compressing dynamic video, because taked the DWT conversion to substitute DCT to the 16x16 luminance block, PSNR can descend especially to the PSNR influence of luminance component obviously (5-10dB), but compression ratio improves big (20%);
2) compare with the current integrated scheme of taking, namely compress dynamically and static 2 kinds of images with standard H.264: the dynamic video aspect is as described in the epimere; Compare with H.264 compressing rest image, under the situation of be more or less the same at PSNR (low about 1dB), the compression ratio of this method be significantly improved (about 14%).Therefore, this method can better meet the application demand of unmanned plane under event of bandwidth limitation.
Claims (2)
1. based on the comprehensive compaction coding method of unmanned plane reconnaissance image of selectivity piece conversion, comprise following step:
(1) macroblock partitions in the scheme H.264;
Coded image is divided by standard H.264, be divided into the macro block of 16 * 16 pixel sizes;
Coded image comprises video information, SAR image, high spectrum image;
(2) according to type selecting intraframe coding or the interframe encoding mode of picture frame, carry out predictive coding respectively;
If the multiframe dynamic video adopts inter prediction and intraframe predictive coding; If the high spectrum image after the SAR image of single frames or the removal spectrum intersexuality then only adopts intraframe predictive coding;
By predictive coding, obtain the predicted value of current macro;
(3) the image residual block calculates;
To subtract each other through macroblock prediction value and the former macro block that step (2) obtain, obtain the image residual block;
(4) based on piece size Selection spatial transform method;
16 * 16 and 8 * 8 luminance pixel pieces are carried out wavelet transform, and all the other macro blocks carry out discrete cosine transform;
(5) entropy coding:
Utilize entropy coding in the standard H.264/AVC to control information and residual error data, motion vector information after quantizing carry out data compression, after finishing the entropy coding, obtain the compressed bit stream of compressed image, use for transmission and storage through the network self-adapting layer.
2. the comprehensive compaction coding method of unmanned plane reconnaissance image based on the conversion of selectivity piece according to claim 1, in described (4), discrete wavelet adopts 9/7 wavelet basis.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310265370.3A CN103347185B (en) | 2013-06-28 | 2013-06-28 | The comprehensive compaction coding method of unmanned plane reconnaissance image based on the conversion of selectivity block |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310265370.3A CN103347185B (en) | 2013-06-28 | 2013-06-28 | The comprehensive compaction coding method of unmanned plane reconnaissance image based on the conversion of selectivity block |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103347185A true CN103347185A (en) | 2013-10-09 |
| CN103347185B CN103347185B (en) | 2016-08-10 |
Family
ID=49281956
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310265370.3A Active CN103347185B (en) | 2013-06-28 | 2013-06-28 | The comprehensive compaction coding method of unmanned plane reconnaissance image based on the conversion of selectivity block |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103347185B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103796011A (en) * | 2014-01-20 | 2014-05-14 | 北京航空航天大学 | Unmanned aerial vehicle reconnaissance image general compression method based on JPEG2000 and interframe compensation |
| CN103796012A (en) * | 2014-01-20 | 2014-05-14 | 北京航空航天大学 | Unmanned aerial vehicle multisource heterogeneous reconnaissance image compression interface adaption method |
| CN104349142A (en) * | 2014-11-03 | 2015-02-11 | 南京航空航天大学 | Layered representation-based unmanned plane video adaptive transmission method |
| WO2017020181A1 (en) * | 2015-07-31 | 2017-02-09 | SZ DJI Technology Co., Ltd. | Method of sensor-assisted rate control |
| CN106506456A (en) * | 2016-10-13 | 2017-03-15 | 重庆金美通信有限责任公司 | A kind of method for ospf protocol message being compressed based on LZW compression algorithms and being encrypted |
| CN109462762A (en) * | 2018-10-26 | 2019-03-12 | 西安科锐盛创新科技有限公司 | Image processing method for unmanned plane |
| US10503175B2 (en) | 2017-10-26 | 2019-12-10 | Ford Global Technologies, Llc | Lidar signal compression |
| US10708617B2 (en) | 2015-07-31 | 2020-07-07 | SZ DJI Technology Co., Ltd. | Methods of modifying search areas |
| WO2022061880A1 (en) * | 2020-09-28 | 2022-03-31 | 深圳市大疆创新科技有限公司 | Movable platform and online data processing method and apparatus therefor, and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001103484A (en) * | 1999-09-29 | 2001-04-13 | Canon Inc | Image processing unit and method therefor |
| US20040062446A1 (en) * | 2002-09-26 | 2004-04-01 | Nec Electronics Corporation | Image processing apparatus supporting both discrete cosine transform and discrete wavelet transform |
| CN1813269A (en) * | 2003-06-30 | 2006-08-02 | 皇家飞利浦电子股份有限公司 | System and method for video processing using overcomplete wavelet coding and circular prediction mapping |
| CN101583032A (en) * | 2009-06-26 | 2009-11-18 | 西安电子科技大学 | Adaptive down-sampling and lapped transform-based image compression method |
| CN103108182A (en) * | 2013-01-18 | 2013-05-15 | 北京航空航天大学 | Multi-source special unmanned plane reconnoitered image general compression method |
-
2013
- 2013-06-28 CN CN201310265370.3A patent/CN103347185B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001103484A (en) * | 1999-09-29 | 2001-04-13 | Canon Inc | Image processing unit and method therefor |
| US20040062446A1 (en) * | 2002-09-26 | 2004-04-01 | Nec Electronics Corporation | Image processing apparatus supporting both discrete cosine transform and discrete wavelet transform |
| CN1813269A (en) * | 2003-06-30 | 2006-08-02 | 皇家飞利浦电子股份有限公司 | System and method for video processing using overcomplete wavelet coding and circular prediction mapping |
| CN101583032A (en) * | 2009-06-26 | 2009-11-18 | 西安电子科技大学 | Adaptive down-sampling and lapped transform-based image compression method |
| CN103108182A (en) * | 2013-01-18 | 2013-05-15 | 北京航空航天大学 | Multi-source special unmanned plane reconnoitered image general compression method |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103796011B (en) * | 2014-01-20 | 2017-04-05 | 北京航空航天大学 | Based on the universal compressed method of unmanned plane reconnaissance image that JPEG2000 and interframe are compensated |
| CN103796012A (en) * | 2014-01-20 | 2014-05-14 | 北京航空航天大学 | Unmanned aerial vehicle multisource heterogeneous reconnaissance image compression interface adaption method |
| CN103796012B (en) * | 2014-01-20 | 2017-02-08 | 北京航空航天大学 | Unmanned aerial vehicle multisource heterogeneous reconnaissance image compression interface adaption method |
| CN103796011A (en) * | 2014-01-20 | 2014-05-14 | 北京航空航天大学 | Unmanned aerial vehicle reconnaissance image general compression method based on JPEG2000 and interframe compensation |
| CN104349142A (en) * | 2014-11-03 | 2015-02-11 | 南京航空航天大学 | Layered representation-based unmanned plane video adaptive transmission method |
| CN104349142B (en) * | 2014-11-03 | 2018-07-06 | 南京航空航天大学 | A kind of UAV Video adaptive transmission method based on layering expression |
| WO2017020181A1 (en) * | 2015-07-31 | 2017-02-09 | SZ DJI Technology Co., Ltd. | Method of sensor-assisted rate control |
| US10708617B2 (en) | 2015-07-31 | 2020-07-07 | SZ DJI Technology Co., Ltd. | Methods of modifying search areas |
| US10834392B2 (en) | 2015-07-31 | 2020-11-10 | SZ DJI Technology Co., Ltd. | Method of sensor-assisted rate control |
| CN106506456A (en) * | 2016-10-13 | 2017-03-15 | 重庆金美通信有限责任公司 | A kind of method for ospf protocol message being compressed based on LZW compression algorithms and being encrypted |
| US10503175B2 (en) | 2017-10-26 | 2019-12-10 | Ford Global Technologies, Llc | Lidar signal compression |
| CN109462762A (en) * | 2018-10-26 | 2019-03-12 | 西安科锐盛创新科技有限公司 | Image processing method for unmanned plane |
| WO2022061880A1 (en) * | 2020-09-28 | 2022-03-31 | 深圳市大疆创新科技有限公司 | Movable platform and online data processing method and apparatus therefor, and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103347185B (en) | 2016-08-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103347185B (en) | The comprehensive compaction coding method of unmanned plane reconnaissance image based on the conversion of selectivity block | |
| CN103108182B (en) | The universal compressed method of multi-source foreign peoples unmanned plane reconnaissance image | |
| CN102369731B (en) | Filter prediction based on activity metrics in video coding | |
| US8396307B2 (en) | Image processing system, image processing method and program | |
| CN103782598A (en) | Fast encoding method for lossless coding | |
| CN102447907A (en) | Video sequence coding method aiming at HEVC (High Efficiency Video Coding) | |
| CN103098463A (en) | Variable length coding of coded block pattern (CBP) in video compression | |
| CN107710758A (en) | HDR and wide color gamut video data are handled for video coding | |
| CN103621082A (en) | Quantization in video coding | |
| CN104041035A (en) | Lossless Coding and Associated Signaling Methods for Compound Video | |
| CN103636207B (en) | For the decoding coefficient decoding of the variable-length of lightness and chrominance block | |
| CN101911702A (en) | Quantization based on rate-distortion modeling for cabac coders | |
| US20110235715A1 (en) | Video coding system and circuit emphasizing visual perception | |
| WO2005122588A1 (en) | Method for chroma deblocking | |
| US20190208229A1 (en) | Image processing system, image processing method and program | |
| US20180302643A1 (en) | Video coding with degradation of residuals | |
| KR20100079037A (en) | Method and apparatus for intra encoding by fast intra mode searching | |
| US8891616B1 (en) | Method and apparatus for entropy encoding based on encoding cost | |
| WO2023040600A1 (en) | Image encoding method and apparatus, image decoding method and apparatus, electronic device, and medium | |
| US20220014744A1 (en) | Adaptive coding of prediction modes using probability distributions | |
| US20100098166A1 (en) | Video coding with compressed reference frames | |
| GB2579753A (en) | Adaptive tile data size coding for video and image compression | |
| WO2009031904A2 (en) | Method for alternating entropy coding | |
| KR100801967B1 (en) | Encoder and decoder for Context-based Adaptive Variable Length Coding, methods for encoding and decoding the same, and a moving picture transmission system using the same | |
| CN109672891B (en) | Lossless secondary compression method of JPEG image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information |
Inventor after: Ding Wenrui Inventor after: Liu Shuo Inventor after: Kang Chuanbo Inventor after: Li Hongguang Inventor after: Lu Aiying Inventor before: Kang Chuanbo Inventor before: Lu Aiying Inventor before: Ding Wenrui Inventor before: Li Hongguang |
|
| COR | Change of bibliographic data | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20170511 Address after: 100191 Haidian District, Xueyuan Road, No. 37, Patentee after: Beijing northern sky long hawk UAV Technology Co. Ltd. Address before: 100191 Haidian District, Xueyuan Road, No. 37, Patentee before: Beihang University |