CN109862207A - A kind of KVM video content change detecting method based on compression domain - Google Patents
A kind of KVM video content change detecting method based on compression domain Download PDFInfo
- Publication number
- CN109862207A CN109862207A CN201910107774.7A CN201910107774A CN109862207A CN 109862207 A CN109862207 A CN 109862207A CN 201910107774 A CN201910107774 A CN 201910107774A CN 109862207 A CN109862207 A CN 109862207A
- Authority
- CN
- China
- Prior art keywords
- frame
- macro block
- content change
- present frame
- present
- 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
Abstract
The present invention relates to a kind of KVM video content change detecting method based on compression domain, comprising the following steps: (1) input the KVM video code flow of an image group；(2) parsing video code flow obtains the compression domain encoded information of a P frame；(3) macro block bit number filters；(4) whether changed using the current content frame of encoded information multi-class discrimination；(5) image group is saved or abandoned according to content change mark.The present invention takes full advantage of the encoded information generated in video coding process, without reconstruction image, the variation of KVM video content can be effectively detected as unit of image group, computation complexity has been greatly reduced in processing method compared to complete pixel domain, while also effectively reducing the storage quantity of KVM video.
Description
Technical field
The present invention relates to Digital Video Processing fields, and in particular to a kind of KVM video content variation inspection based on compression domain
Survey method.
Background technique
The long-range handover management system of keyboard (Keyboard), display (Video) and mouse (Mouse), i.e. kvm system,
Also referred to as multicomputer control system is mainly used for various computer remote controls in computer room.KVM video is digital KVM
Compressed data are acquired to the original video of computer screen output in system, in recent years, with high-performance computer, height
The winged degree of broadband network develops, and the magnanimity KVM video that the application of kvm system is continuously increased, and saves is to storage and management band
Big inconvenience is carried out.In order to reduce the amount of storage of KVM video and facilitate management, the variation of KVM video content is detected very
It is important.Existing in KVM video much can be only right by carrying out content change detection to it without the redundancy section of content change
The part that video content changes is stored, and is abandoned to the unconverted redundancy section of video content, is deposited to reduce
Reserves, while also can be convenient the data management in later period.In addition, the variation detection of KVM video content is answered in some specific areas
With also than wide, than an alarm pop-up may be will pop up at work if any some industrial control equipments, become using content
The method for changing detection can detecte this alarm pop-up, and handle this alarm pop-up in time, it is possible to reduce some unnecessary
Loss.
The KVM video of magnanimity is encoded and is transmitted by the H.264 standard of mainstream mostly at present.In order to reduce
Kvm system delay, in h .264, mainly I frame and P frame are used for coding transmission, and I frame is also known as intracoded frame, are a kind of included
The independent frame of all information can be independently decoded without reference to other images, and P frame is also known as inter prediction encoding frame, needs
With reference to it is preceding to reconstructed frame just can be carried out coding.H.264 using the macro block of 16 × 16 sizes, the volume of macro block in compression standard
Code type is divided into Intra macro block, Inter macro block.Intra macro block is predicted by reconstructed data in present frame, and Inter
Macro block is predicted that motion vector refers to that current macro is inclined with respect to reference frame image using the reference frame image rebuild
It moves, includes horizontal component and vertical component.
Subtractive method, frame difference method, background modeling method, base currently, existing video content change detection techniques mainly have powerful connections
Motion detection in feature, the motion detection based on optic flow technique.Application No. is the patents of CN201210334876 to disclose one
The method of kind computer output video image variation detection.This method changes into gray level image first with by color image；It again will be grey
It spends image and carries out mean filter；Then filtered image is subjected to poor piecemeal, frame, mean value and variance；Finally according to mean value, side
Poor computational discrimination threshold value, to judge whether to change.Application No. is the patents of CN201110009716 to disclose a kind of base
In the image change detection method of principal component broad sense inverse transformation.This method after reorganizing data to image to be detected by carrying out
Transformation in two feature spaces, and be updated wave band difference processing in feature space after the conversion and be changing into obtaining
Point, then region of variation is extracted by automatic threshold method, realize Image Change Detection.Application No. is CN200810101240.5
Patent disclose a kind of method detected according to scene variation characteristic to moving objects in video.This method is first according to training
Image carries out characteristic point detection, background characteristics point parameter and model pixel point parameter and calculates, and using background characteristics point parameter and works as
The image characteristic point of preceding detection calculates motion feature point；Using motion feature point to pixel parameter minute in background model
Class；Background model pixel parameter renewal rate is calculated using classification results；Finally according to present image and background model pixel
Point parameter, obtains the location and shape of moving object using background difference method.
The above method mainly carries out video content variation detection in pixel domain, and overall calculation complexity is higher, furthermore right
It in KVM video, needs to carry out the decoding and rebuilding of pixel domain, further increases calculation amount, therefore KVM cannot be suitable for well
The content change of video detects.In H.264 code stream, frame bit number, the macro block (mb) type of reflection content change are inherently contained
It with the encoded informations such as motion vector, can only build without being finished full weight a parsing code stream, therefore KVM is regarded in H.264 compression domain
Frequency, which carries out content change detection, can guarantee lower computation complexity.
Summary of the invention
In order to reduce the data volume of KVM video code flow preservation, the present invention utilizes the parsing from code stream to obtain encoded information, mentions
A kind of method for having supplied KVM video content variation detection based on compression domain, the described method comprises the following steps:
(1) the KVM video code flow of an image group is inputted
The KVM video code flow of input is that H.264 code stream, the frame type of support are I frame (intracoded frame) and P frame (forward direction
Inter-frame encoding frame), an image group (GOP) is made of an I frame and subsequent multiple P frames, P frame without force macro block updating,
That is encoder does not force the type of coding of specified macro block, but preferably obtains from various type of codings, to make the macro block of P frame
Type of coding is able to reflect out its picture material difference degree with former frame.
(2) parsing video code flow obtains the compression domain encoded information of a P frame
Parse the type of coding, bit number and motion vector that obtained P frame encoded information includes all macro blocks of the frame；If should
Frame is first P frame, then setting content change mark CF is 0, and parses a P frame again as present frame；The type of coding of macro block
Including Intra macro block (inter-coded macroblocks) and Inter macro block (inter-coded macroblocks), according to statistics, the Intra in P frame is macro
Block often has in the region of video content significant changes, and other situations are mostly Inter macro block.I frame is due to no inter prediction encoding
Information, only Intra macro block, do not have Inter macro block, and the picture material difference journey of before and after frames cannot be predicted by macro block (mb) type
Degree, therefore the present invention is not handled it.
(3) macro block bit number filters
Current most of KVM video code flows are all to encode to obtain after VGA mouthfuls of analog-to-digital conversions, and the video after coding exists
Some noises influence the detection of its content change, it is therefore desirable to be filtered to encoded information.It is complete by analysis code Stream Discovery
Complete static macro block bit number is equal or close to zero, and its bit number of the obvious macro block of content change is larger, there is part and makes an uproar
Its bit number non-zero of the static macro block of sound and smaller.According to these features, the present invention is filtered macro block bit number: will be less than
Threshold value THbit1Macro block bit number be all directly set to zero, by analysis statistics KVM video code flow in there are the macro block bits of noise
Number obtains THbit1Value range be [50,300], user chooses according to the noise situations of practical code stream, the big value of noise
Greatly.
(4) whether changed using the current content frame of encoded information multi-class discrimination
Using the encoded informations such as present frame and the macroblock encoding type of former frame, macro block bit number and motion vector, for
It is unchanged that KVM video typical scene differentiates that video content has, the specific steps are as follows:
The differentiation whether (4-1) integrally has significant change to desktop
If the Intra macro block number of present frame is greater than the Intra macro block number of former frame, and the Intra of present frame is macro
The percentage of the total macro block number of block number Zhan is greater than threshold value THintra1, then illustrate present frame in terms of content compared to former frame have compared with
Big variation, setting content change mark CF is 1, and skips to step (5), otherwise carries out next step detection；By statisticalling analyze desktop
The Intra macro block number of P frame, obtains TH in the scene that integrally has significant changeintra1Value range be [40,90].
Whether (4-2) occurs the differentiation of operation box variation to computer desktop
If the Intra macro block number of present frame is greater than the Intra macro block number of former frame and difference is greater than threshold value THintra2,
And the maximum Intra macro block number for having four neighborhoods to be connected in present frame is greater than threshold value THintra3, then illustrate exist in present frame
Picture material significant change occurs for some region, and setting content change mark CF is 1, skips to step (5), otherwise carries out next step inspection
It surveys；There is the Intra macro block of Intra macro block number and its former frame in the P frame of operation box by statisticalling analyze computer desktop
The maximum Intra macro block number of four neighborhoods connection obtains TH in the several and frameintra2Value range be [20,1000],
THintra3Value range be [3,300], chosen according to the image resolution ratio and operation box size that are actually subjected to detection, image point
Resolution is bigger, object run frame is bigger, and value is bigger.The four neighborhoods connection is current macro and upper macroblock, lower section
Macro block, left side macro block or the connection of the right macro block, are a kind of mode of communicating of classics in field of image processing.
The differentiation of (4-3) to minor changes such as mouse and characters
If there are bit numbers to be all larger than threshold value TH for present framebit2Four neighborhoods connection macro block and the macro block number be greater than threshold
Value THmb1Or the bit number that present frame at least has 1 Intra macro block is greater than threshold value THbit2, then in both cases all
Illustrating that picture material significant changes have occurred compared with former frame in discreet region for present frame, the bit number of corresponding macro block is especially big,
Therefore setting content change mark CF is 1, skips to step (5), otherwise carries out next step detection；Have by statisticalling analyze computer desktop
Macro block bit number and its four neighborhood adjacent macroblocks in the mobile P frame with character input of mouse obtain THbit2Value range be
[200,1000]、THmb1Value range be [1,4].
(4-4) whether there is or not the differentiations of the variations such as overall movement to application software
The motion vector magnitude MV of each macro block of present frame is calculated by formula (1) firstap, wherein | MVx| and | MVy| it is
The horizontal component amplitude and vertical component amplitude of the motion vector of the obtained each macro block of present frame are parsed according to step (2)；If
There are MV in present frameapIt is all larger than threshold value THmvFour neighborhoods connection macro block and the macro block number be greater than threshold value THmb2, then say
The content in bright current frame portion region sends the variation moved integrally, and motion vector magnitude is larger, sets content change mark CF and is
1；By statisticalling analyze, there are the biggish macroblock motion vector of amplitude and its neighbours in the P frame of the variations such as application software overall movement
Domain adjacent macroblocks obtain THmvValue range be [16,128], THmb2Value range be [4,100]；Intra macro block without motion
Vector, MVapAll it is set as 0.
MVap=| MVx|+|MVy| (1)
(5) image group code stream is saved or abandoned according to content change mark
If the content change mark CF of present frame is 1, illustrate that the content of an at least frame in present image group changes,
Code stream analyzing and the differentiation for then no longer carrying out subsequent frame, save present image group video code flow；If the content change mark of present frame
Will CF is 0 and present frame is the last one P frame of present image group, then abandons present image group video code flow；If present frame
Content change mark CF is 0 and present frame is not the last one P frame of present image group, then repeatedly at step (2) to step (4)
Manage the detection of next P frame.
Compared with prior art, the invention has the following advantages:
The variation detection of KVM video content directly is carried out using the compression domain encoded information for including in H.264 code stream, without weight
Image is built, can rapidly detection content be changed as unit of image group, while also effectively reducing the storing data of KVM video
Computation complexity has been greatly reduced in amount, the detection method compared to pixel domain.
Detailed description of the invention
Fig. 1 is the flow chart of the method for the present invention.
Specific embodiment
Carry out the present invention is described in detail below with reference to embodiment and attached drawing, but the present invention is not limited to this.The present embodiment is adopted
Video Decoder is FFmpeg, version number 3.2.4；The video source of input is the view that KVM switcher is recorded from VGA mouthfuls
Frequently, resolution ratio 1080P, the target bit rate of recording are 4Mbps, frame per second 30fps, and encoding pre- geodesic structure is IPPP mode, I frame
Between be divided into 150；The KVM video content of recording include mouse movement, pop-up, input Chinese character, sliding word document, desktop integrally compared with
The scenes such as fast large change.As shown in Figure 1, a kind of KVM video content change detecting method based on compression domain, including following step
It is rapid:
(1) the KVM video code flow of an image group is inputted；
(2) parsing video code flow obtains the compression domain encoded information of a P frame；
(3) macro block bit number filters；
(4) whether changed using the current content frame of encoded information multi-class discrimination；
(5) image group is saved or abandoned according to content change mark.
Step (1) specifically includes:
An image group video code flow is inputted, which is H.264 code stream, and first is I frame and subsequent
149 P frames.
Step (2) specifically includes:
Parse the type of coding, bit number and motion vector that obtained P frame encoded information includes all macro blocks of the frame；If should
Frame is first P frame, then setting content change mark CF is 0, and parses a P frame again as present frame；The type of coding of macro block
Including Intra macro block and Inter macro block.The present invention is not handled I frame.
Step (3) specifically includes:
The size for counting bit number shared by each macro block in present frame, will be less than threshold value THbit1Macro block bit number it is direct
It is set to zero；Wherein THbit1Value range be [50,300], take 200 herein.
Step (4) specifically includes:
Using the encoded informations such as present frame and the macroblock encoding type of former frame, macro block bit number and motion vector, for
It is unchanged that KVM video typical scene differentiates that video content has, and specifically includes:
The differentiation whether (4-1) integrally has significant change to desktop
If the Intra macro block number of present frame is greater than the Intra macro block number of former frame, and the Intra of present frame is macro
The percentage of the total macro block number of block number Zhan is greater than threshold value THintra1, then setting content change mark CF is 1, and skips to step (5),
Otherwise next step detection is carried out；Wherein THintra1Value range be [40,90], take 70 herein.
Whether (4-2) occurs the differentiation of operation box variation to computer desktop
If the Intra macro block number of present frame is greater than the Intra macro block number of former frame and difference is greater than threshold value THintra2,
And the maximum Intra macro block number for having four neighborhoods to be connected in present frame is greater than threshold value THintra3, then content change mark CF is set
It is 1, skips to step (5), otherwise carries out next step detection；Wherein THintra2Value range be [20,1000], take 80 herein,
THintra3Value range be [3,300], take 20 herein.
The differentiation of (4-3) to minor changes such as mouse and characters
If there are bit numbers to be all larger than threshold value TH for present framebit2Four neighborhoods connection macro block and its macro block number be greater than threshold value
THmb1Or the bit number that present frame at least has 1 Intra macro block is greater than threshold value THbit2, then all set in both cases
Content change mark CF is 1, skips to step (5), otherwise carries out next step detection；Wherein THbit2Value range be [200,
1000], 300, TH are taken hereinmb1Value range be [Isosorbide-5-Nitrae], take 2 herein.
(4-4) whether there is or not the differentiations of the variations such as overall movement to application software
The motion vector magnitude MVa of each macro block of present frame is calculated by formula (1) firstp, wherein | MVx| and | MVy| it is
The horizontal component amplitude and vertical component amplitude of the motion vector of the obtained each macro block of present frame are parsed according to step (2)；If
There are MV in present frameapIt is all larger than threshold value THmvFour neighborhoods connection macro block and the macro block number be greater than threshold value THmb2, then set
Content change mark CF is 1；Wherein THmvValue range be [16,128], take 50, TH hereinmb2Value range be [4,
100], 10 are taken herein；Intra macro block without motion vector, MVapAll it is set as 0.
MVap=| MVx|+|MVy| (1)
Step (5) specifically includes:
If the content change mark CF of present frame is 1, illustrate that the content of an at least frame in present image group changes,
Code stream analyzing and the differentiation for then no longer carrying out subsequent frame, save present image group video code flow；If the content change mark of present frame
Will CF is 0 and present frame is the last one P frame of present image group, then abandons present image group video code flow；If present frame
Content change mark CF is 0 and present frame is not the last one P frame of present image group, then repeatedly at step (2) to step (4)
Manage the detection of next P frame.
Claims (3)
1. a kind of KVM video content change detecting method based on compression domain, which is characterized in that the detection method include with
Lower step:
(1) the KVM video code flow of an image group is inputted:
The KVM video code flow of input be H.264 code stream, an image group be made of an I frame and subsequent multiple P frames, P frame not into
Row forces macro block updating；
(2) parsing video code flow obtains the compression domain encoded information of a P frame:
The compression domain encoded information for parsing obtained P frame includes the type of coding, bit number and motion vector of all macro blocks of the frame；
If the frame is first P frame, setting content change mark CF is 0, and parses a P frame again as present frame；The coding of macro block
Type includes inter-coded macroblocks Intra macro block and inter-coded macroblocks Inter macro block；
(3) macro block bit number filters:
The size for counting bit number shared by each macro block in present frame, will be less than threshold value THbit1Macro block bit number be all directly set to
Zero；
(4) whether changed using the current content frame of encoded information multi-class discrimination:
Using present frame and the macroblock encoding type of former frame, bit number and motion vector, differentiate that whether there is or not changes for KVM video content
Change；
(5) it saves according to content change mark or abandons image group:
If the content change mark CF of present frame is 1, the no longer code stream analyzing of progress subsequent frame and differentiation saves present image
Group video code flow；If the content change mark CF of present frame is 0 and present frame is the last one P frame of present image group, lose
Abandon present image group video code flow；If the content change mark CF of present frame is 0 and present frame is not the last of present image group
One P frame, then repeatedly step (2) handles the detection of next P frame to step (4).
2. a kind of KVM video content change detecting method based on compression domain as described in claim 1, which is characterized in that institute
It states in step (3), THbit1Value range be [50,300].
3. a kind of KVM video content change detecting method based on compression domain as described in claim 1, which is characterized in that institute
Include following sub-step in the step of stating (4):
(4-1) is if the Intra macro block number of present frame is greater than the Intra macro block number of former frame, and the Intra of present frame is macro
The percentage of the total macro block number of block number Zhan is greater than threshold value THintra1, then setting content change mark CF is 1, is jumped to step (5), no
Then carry out next step detection；Wherein THintra1Value range be [40,90]；
(4-2) is if the Intra macro block number of present frame is greater than the Intra macro block number of former frame and difference is greater than threshold value
THintra2, and the maximum Intra macro block number for having four neighborhoods to be connected in present frame is greater than threshold value THintra3, then content change is set
Indicate that CF is 1, skips to step (5), otherwise carry out next step detection；Wherein THintra2Value range be [20,1000],
THintra3Value range be [3,300]；
(4-3) is if there are bit numbers to be all larger than threshold value TH for present framebit2Four neighborhoods connection macro block and the number of the macro block be greater than
Threshold value THmb1Or the bit number that present frame at least has 1 Intra macro block is greater than threshold value THbit2, then content change mark is set
CF is 1, skips to step (5), otherwise carries out next step detection；Wherein THmb1Value range be [1,4], THbit2Value model
It encloses for [200,1000]；
The motion vector magnitude MV of each macro block of present frame is calculated in (4-4) by formula (1) firstap, wherein | MVx| and | MVy| it is
The horizontal component amplitude and vertical component amplitude of the motion vector of the obtained each macro block of present frame are parsed according to step (2),
MVap=| MVx|+|MVy| (1)
If there are MV in present frameapIt is all larger than threshold value THmvFour neighborhoods connection macro block and the macro block number be greater than threshold value
THmb2, then setting content change mark CF is 1；Wherein THmvValue range be [16,128], THmb2Value range be [4,
100]。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910107774.7A CN109862207B (en) | 2019-02-02 | 2019-02-02 | KVM video content change detection method based on compressed domain |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910107774.7A CN109862207B (en) | 2019-02-02 | 2019-02-02 | KVM video content change detection method based on compressed domain |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109862207A true CN109862207A (en) | 2019-06-07 |
| CN109862207B CN109862207B (en) | 2021-04-09 |
Family
ID=66897566
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910107774.7A Active CN109862207B (en) | 2019-02-02 | 2019-02-02 | KVM video content change detection method based on compressed domain |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109862207B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112637538A (en) * | 2020-12-15 | 2021-04-09 | 重庆紫光华山智安科技有限公司 | Smart tag method, system, medium, and terminal for optimizing video analysis |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060109902A1 (en) * | 2004-11-19 | 2006-05-25 | Nokia Corporation | Compressed domain temporal segmentation of video sequences |
| CN101090498A (en) * | 2007-07-19 | 2007-12-19 | 华为技术有限公司 | Device and method for motion detection of image |
| CN101246547A (en) * | 2008-03-03 | 2008-08-20 | 北京航空航天大学 | Method for detecting moving objects in video according to scene variation characteristic |
| CN102063722A (en) * | 2011-01-18 | 2011-05-18 | 上海交通大学 | Image change detecting method based on principle component general inverse transformation |
| CN102867308A (en) * | 2012-09-11 | 2013-01-09 | 大连理工大学 | Method for detecting change of video image output by computer |
| CN103985250A (en) * | 2014-04-04 | 2014-08-13 | 浙江工业大学 | Light-weight holographic road traffic state visual inspection device |
| CN104683802A (en) * | 2015-03-24 | 2015-06-03 | 江南大学 | H.264/AVC compressed domain based moving target tracking method |
| CN109104609A (en) * | 2018-09-12 | 2018-12-28 | 浙江工业大学 | A kind of lens boundary detection method merging HEVC compression domain and pixel domain |
-
2019
- 2019-02-02 CN CN201910107774.7A patent/CN109862207B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060109902A1 (en) * | 2004-11-19 | 2006-05-25 | Nokia Corporation | Compressed domain temporal segmentation of video sequences |
| CN101090498A (en) * | 2007-07-19 | 2007-12-19 | 华为技术有限公司 | Device and method for motion detection of image |
| CN101246547A (en) * | 2008-03-03 | 2008-08-20 | 北京航空航天大学 | Method for detecting moving objects in video according to scene variation characteristic |
| CN102063722A (en) * | 2011-01-18 | 2011-05-18 | 上海交通大学 | Image change detecting method based on principle component general inverse transformation |
| CN102867308A (en) * | 2012-09-11 | 2013-01-09 | 大连理工大学 | Method for detecting change of video image output by computer |
| CN103985250A (en) * | 2014-04-04 | 2014-08-13 | 浙江工业大学 | Light-weight holographic road traffic state visual inspection device |
| CN104683802A (en) * | 2015-03-24 | 2015-06-03 | 江南大学 | H.264/AVC compressed domain based moving target tracking method |
| CN109104609A (en) * | 2018-09-12 | 2018-12-28 | 浙江工业大学 | A kind of lens boundary detection method merging HEVC compression domain and pixel domain |
Non-Patent Citations (1)
| Title |
|---|
| 杨洋 等: "《一种HEVC压缩域的运动目标检测方法》", 《小型微型计算机系统》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112637538A (en) * | 2020-12-15 | 2021-04-09 | 重庆紫光华山智安科技有限公司 | Smart tag method, system, medium, and terminal for optimizing video analysis |
| CN112637538B (en) * | 2020-12-15 | 2022-08-02 | 重庆紫光华山智安科技有限公司 | Smart tag method, system, medium, and terminal for optimizing video analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109862207B (en) | 2021-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5873536B2 (en) | Dynamic selection of motion prediction search range and extended motion vector range | |
| KR101599875B1 (en) | Method and apparatus for multimedia encoding based on attribute of multimedia content, method and apparatus for multimedia decoding based on attributes of multimedia content | |
| JP5767388B2 (en) | Video encoding method and apparatus, video decoding method and apparatus | |
| US10003792B2 (en) | Video encoder for images | |
| US7606310B1 (en) | Video compression and encoding method | |
| GB2531086A (en) | Method and apparatus for vector encoding in video coding and decoding | |
| JP2000224590A (en) | Method for extracting characteristics of video sequence | |
| CN105847811A (en) | Method and apparatus for entropy encoding a transform block, and to a method and apparatus for entropy decoding the transform block | |
| KR102638065B1 (en) | Methods and apparatus for multiple line intra prediction in video compression | |
| KR20240038959A (en) | Method and system for performing combined inter and intra prediction | |
| CN114731440A (en) | Lossless encoding of video data | |
| CN102934445A (en) | Methods and apparatuses for encoding and decoding image based on segments | |
| CN109862207A (en) | A kind of KVM video content change detecting method based on compression domain | |
| CN113079373A (en) | Video coding method based on HEVC-SCC | |
| JPH08298665A (en) | Circuit and method for detecting motion vector | |
| Bachu et al. | Adaptive order search and tangent-weighted trade-off for motion estimation in H. 264 | |
| EP3926584A1 (en) | Method, computer program and system for detecting changes and moving objects in a video view | |
| KR20230115043A (en) | Video processing method and video processing apparatus using super resolution deep learning network based on image quality | |
| KR20230003491A (en) | Methods and systems of video coding using reference regions | |
| CN1338235A (en) | Method and device for video compress encoding based on division | |
| US20240121395A1 (en) | Methods and non-transitory computer readable storage medium for pre-analysis based resampling compression for machine vision | |
| Mao et al. | A fast intra prediction algorithm based on wmse for 360-degree video | |
| US20240114185A1 (en) | Video coding for machines (vcm) encoder and decoder for combined lossless and lossy encoding | |
| US20240048709A1 (en) | Methods and systems for temporal resampling for multi-task machine vision | |
| US20240121445A1 (en) | Pre-analysis based image compression methods |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |