CN104618679A - Method for extracting key information frame from monitoring video - Google Patents
Method for extracting key information frame from monitoring video Download PDFInfo
- Publication number
- CN104618679A CN104618679A CN201510062263.XA CN201510062263A CN104618679A CN 104618679 A CN104618679 A CN 104618679A CN 201510062263 A CN201510062263 A CN 201510062263A CN 104618679 A CN104618679 A CN 104618679A
- Authority
- CN
- China
- Prior art keywords
- frame
- timeslice
- active
- time
- video
- 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 invention discloses a method for extracting key information frame from a monitoring video. The method is that the video is divided into three types, namely, moving time piece, static time piece, and discontinuous moving time piece; the moving time piece contains more usable information and the frame can be extracted at relatively small intervals; the static time piece contains less usable information and the frame can be extracted at large intervals; the discontinuous moving time piece possibly contains important information and the frame containing the moving objects is pumped. According to the core technology, the monitoring video is divided into the time pieces of the three types, different frame extracting strategies are performed for the pieces of different types, and therefore, more usable information can remained while the quantity of the monitoring video frames is greatly decreased.
Description
Technical field
The present invention relates to field of computer technology, particularly relate to a kind of method extracting key message frame in monitor video.
Background technology
Current era, video monitoring is ubiquitous.Monitor video has become the important power-assisted of the delinquent one's share of expenses for a joint undertaking of fright and public security organ's clear up a criminal case.But the monitor video of magnanimity is while offering convenience to cracking of cases, also bring huge difficulty to video storage management and searching of relevant information.Consider that a lot of monitor video includes a large amount of time slices without personnel activity, the amount of information useful to detection that these time slices comprise is less, if automatically can identify and filter out these time slices, greatly can save video storage space and shorten the time that detection personnel check video.
In the past few years, the technology extracting Video Key information frame has had a lot of research work.Video concentration technique is shown by the moving object of different time fragment in same monitor video scene is put into fragment at the same time, bulk information can be shown at short notice, and can find that such as someone frequently occurs (offender sets foot-point) in certain place some special situations.But these class methods more pay close attention to background modeling, moving object is extracted, and moving object concentrates the research of strategy, finally more chaotic in the concentrated video clips scene formed, and brings difficulty to checking of detection personnel.Also have research to propose from different perspectives and extract the method for monitoring video frame, such as Yuan is little happyly proposes the method extracting frame of video in the safety time section of setting, and the method for frame rate is taken out in Pan Guoping backup harddisk capacity setting of giving chapter and verse.But they have done indifference process to all monitoring video frame, and for cracking of cases, the useful information having the frame of personnel activity to comprise than the frame without personnel activity is much bigger, should preferentially retain; Frame without personnel activity then can be rejected in a large number.
Summary of the invention
The object of the present invention is to provide a kind of method extracting key message frame in monitor video, automatically can identify and filter out time slice useless in monitor video.
In order to reach above technique effect, technical scheme of the present invention is as follows:
Extract a method for key message frame in monitor video, comprise the following steps:
S1, mark active frame/frozen frozen mass, the 1st frame of monitor video is not marked, and from the 2nd frame, the luminance difference according to present frame and former frame marks it for active frame or frozen frozen mass.
S2, time division sheet, be divided into three types video segment by monitor video:
Activity time sheet (P1): each the frame video in timeslice is all marked as active frame and timeslice reaches certain length;
Sheet quiescent time (P2): each the frame video in timeslice is all marked as frozen frozen mass and timeslice reaches certain length;
Intermittently active timeslice (P3): both containing active frame in timeslice, also containing frozen frozen mass, but same frame of video continuous print quantity do not reach corresponding activity time sheet or quiescent time sheet time span requirement.
S3, extraction frame of video, activity time sheet includes more useful information, can take out frame with less interval; The useful information that quiescent time comprises in sheet is little, can take out frame by large-spacing; Two kinds of situations then can be further divided in intermittently active timeslice, one is that active frame is little, the frequency of occurrences is very low, in this case the larger probability being labeled as active frame is because the reasons such as light change cause, instead of real have action figure, so can by quiescent time sheet frame strategy of taking out extract; Another kind of situation is that active frame is more, and frequently occur, then active frame has and may have real moving object greatlyr, adjacent frozen frozen mass may be then can not see by monitored camera after moving object appears at shelter, in this case take out frame with regard to needing by less interval, and extract the motion frame in intermittently active timeslice as far as possible.
Preferably, the computational methods marking active frame/frozen frozen mass are as follows:
Wherein, I
i, jt () is the pixel brightness value of t frame position (i, j), I
i, j(t-1) be the pixel brightness value of t-1 frame position (i, j), d (t) represents the absolute value sum of t frame and its former frame correspondence position pixel intensity difference; T1 is the threshold value that user sets, and detects the size of moving object as required, can suitably change.In common application, T1 is set to 1/4 of brightness maxima, and such as 8-bit video T1 can be set to 64.When the luminance difference of t frame and former frame exceedes threshold value T1, t frame is just marked as 1 (i.e. active frame), otherwise is labeled as 0 (i.e. frozen frozen mass).
Preferably, the method for time division sheet is as follows:
Wherein, p refers to timeslice type, and N{b (t)=1} represents the quantity of continuously active frame, N{b (t)=0} represents the quantity of continuous frozen frozen mass, and T2, T3 are the threshold values that user sets, according to concrete monitoring scene and task needs, this threshold value is arranged can certain change.The basic principle of setting is that, when video primary dcreening operation, T2 establishes height, and T3 establishes height; And need to some have a mind to avoid the criminal of video monitoring do meticulous detection time, then T2 establishes low, and T3 establishes low; T2, T3 are set as the frame number of 1 second under normal circumstances.
Preferably, the method extracting frame of video is as follows:
Activity time sheet (P1): take out frame by comparatively small time intervals T4, wherein T4 is user's setting threshold, advises choosing value within 0.5-2 second, is set as 1 second under normal circumstances.
Sheet quiescent time (P2): take out frame by larger time interval T5, wherein T5 is user's setting threshold, advises value more than 2 seconds, can just get 2 frames in sheet a quiescent time under normal circumstances, namely timeslice start and at the end of each frame.Intermittently active timeslice (P3): be divided into two kinds of situations, one be in timeslice continuously active frame time length be less than 0.2 second and the ratio of active frame in timeslice lower than 20% time, take out frame by larger time interval T5, wherein T5 is user's setting threshold, advises value more than 2 seconds; Otherwise take out frame by comparatively small time intervals T6, wherein T6 is user's setting threshold, advises value more than 1 second.With activity time sheet and quiescent time sheet take out frame method unlike, be not that fixed intervals take out frame to the frame of taking out in intermittently active timeslice, but extract be greater than setting threshold after first active frame.Such as: to video sequence below, wherein the 1st, 2,21,22,41-44,69, the frame of video such as 70 are marked as active frame, other are labeled as frozen frozen mass.If monitor video frame per second is 25 frames/second, the frame number that T2, T3 are set as 1 second that is 25, then the continuously active frame number in this timeslice or continuous static frame number all not more than 25, so this timeslice is divided into intermittently active timeslice.And due to continuously active frame number maximum quantity be 4 (i.e. 41-44), be less than 0.2 second duration and active frame ratio is 10/71, be less than 20%, therefore take out frame by the first situation of intermittently active timeslice and T5 interval, the frame number being set as 2 seconds durations that is 50.Last frame result of taking out is the 1st frame, and first active frame i.e. the 69th frame after the 50th frame.
| Video sequence | 1 | 1 | 0…0 | 1 | 1 | 0…0 | 1…1 | 0…0 | 1 | 1 |
| Frame number | 1 | 2 | 3-20 | 21 | 22 | 23-40 | 41-44 | 45-68 | 69 | 70 |
| Extract frame | √ | √ |
The beneficial effect adopting technique scheme to bring is: video is divided into " activity time sheet " by the present invention, " sheet quiescent time ", and " intermittently active timeslice " three types, in this three types, activity time sheet includes more useful information, can take out frame with less interval; The useful information that quiescent time comprises in sheet is little, can take out frame by large-spacing; And important information in intermittently active timeslice, may be included, take out frame and will extract frame containing moving object as far as possible.The core technology of this patent is time slice monitor video being divided into this three types, then uses dissimilar fragment and different takes out frame strategy, thus can while greatly shortening monitor video number of frames reservation useful information as much as possible.
Accompanying drawing explanation
Fig. 1 is for taking out frame schematic flow sheet.
Embodiment
Describe technical scheme of the present invention in detail below in conjunction with specific embodiment, be used for explaining technical scheme of the present invention in this illustrative examples of the present invention and explanation, but not as a limitation of the invention.
Embodiment 1:
Extract a method for key message frame in monitor video, comprise the following steps:
Step one, mark active frame/frozen frozen mass, the 1st frame of monitor video is not marked, and from the 2nd frame, the luminance difference according to present frame and former frame marks it for active frame or frozen frozen mass, and computational methods are as follows:
Wherein, I
i, jt () is the pixel brightness value of t frame position (i, j), I
i, j(t-1) be the pixel brightness value of t-1 frame position (i, j), d (t) represents the absolute value sum of t frame and its former frame correspondence position pixel intensity difference; T1 is the threshold value that user sets, and detects the size of moving object as required, can suitably change.In common application, T1 is set to 1/4 of brightness maxima, and such as 8-bit video T1 can be set to 64.When the luminance difference of t frame and former frame exceedes threshold value T1, t frame is just marked as 1 (i.e. active frame), otherwise is labeled as 0 (i.e. frozen frozen mass).
Step 2, time division sheet, is divided into three types video segment by monitor video:
Activity time sheet (P1): each the frame video in timeslice is all marked as active frame and timeslice reaches certain length; Sheet quiescent time (P2): each the frame video in timeslice is all marked as frozen frozen mass and timeslice reaches certain length;
Intermittently active timeslice (P3): both containing active frame in timeslice, also containing frozen frozen mass, but same frame of video continuous print quantity do not reach corresponding activity time sheet or quiescent time sheet time span requirement.
The method divided is as follows:
Wherein, p refers to timeslice type, and N{b (t)=1} represents the quantity of continuously active frame, N{b (t)=0} represents the quantity of continuous frozen frozen mass, and T2, T3 are the threshold values that user sets, according to concrete monitoring scene and task needs, this threshold value is arranged can certain change.The basic principle of setting is that, when video primary dcreening operation, T2 establishes height, and T3 establishes height; And need to some have a mind to avoid the criminal of video monitoring do meticulous detection time, then T2 establishes low, and T3 establishes low; T2, T3 are set as the frame number of 1 second under normal circumstances.
Step 3, extract frame of video, activity time sheet includes more useful information, can take out frame with less interval; The useful information that quiescent time comprises in sheet is little, can take out frame by large-spacing; Two kinds of situations then can be further divided in intermittently active timeslice, one is that active frame is little, the frequency of occurrences is very low, in this case the larger probability being labeled as active frame is because the reasons such as light change cause, instead of real have action figure, so can by quiescent time sheet frame strategy of taking out extract; Another kind of situation is that active frame is more, and frequently occur, then active frame has and may have real moving object greatlyr, adjacent frozen frozen mass may be then can not see by monitored camera after moving object appears at shelter, in this case take out frame with regard to needing by less interval, and extract the motion frame in intermittently active timeslice as far as possible.
The method extracting frame of video is as follows:
Activity time sheet (P1): take out frame by comparatively small time intervals T4, wherein T4 is user's setting threshold, advises choosing value within 0.5-2 second, is set as 1 second under normal circumstances.
Sheet quiescent time (P2): take out frame by larger time interval T5, wherein T5 is user's setting threshold, advises value more than 2 seconds, can just get 2 frames in sheet a quiescent time under normal circumstances, namely timeslice start and at the end of each frame.
Intermittently active timeslice (P3): be divided into two kinds of situations, one be in timeslice continuously active frame time length be less than 0.2 second and the ratio of active frame in timeslice lower than 20% time, take out frame by larger time interval T5, wherein T5 is user's setting threshold, advises value more than 2 seconds; Otherwise take out frame by comparatively small time intervals T6, wherein T6 is user's setting threshold, advises value more than 1 second.With activity time sheet and quiescent time sheet take out frame method unlike, be not that fixed intervals take out frame to the frame of taking out in intermittently active timeslice, but extract be greater than setting threshold after first active frame.Such as: to video sequence below, wherein the 1st, 2,21,22,41-44,69, the frame of video such as 70 are marked as active frame, other are labeled as frozen frozen mass.If monitor video frame per second is 25 frames/second, the frame number that T2, T3 are set as 1 second that is 25, then the continuously active frame number in this timeslice or continuous static frame number all not more than 25, so this timeslice is divided into intermittently active timeslice.And due to continuously active frame number maximum quantity be 4 (i.e. 41-44), be less than 0.2 second duration and active frame ratio is 10/71, be less than 20%, therefore take out frame by the first situation of intermittently active timeslice and T5 interval, the frame number being set as 2 seconds durations that is 50.Last frame result of taking out is the 1st frame, and first active frame i.e. the 69th frame after the 50th frame.
| Video sequence | 1 | 1 | 0…0 | 1 | 1 | 0…0 | 1…1 | 0…0 | 1 | 1 |
| Frame number | 1 | 2 | 3-20 | 21 | 22 | 23-40 | 41-44 | 45-68 | 69 | 70 |
| Extract frame | √ | √ |
Video is divided into " activity time sheet " by the present invention, " sheet quiescent time ", and " intermittently active timeslice " three types, and in this three types, activity time sheet includes more useful information, can take out frame with less interval; The useful information that quiescent time comprises in sheet is little, can take out frame by large-spacing; And important information in intermittently active timeslice, may be included, take out frame and will extract frame containing moving object as far as possible.The core technology of this patent is time slice monitor video being divided into this three types, then uses dissimilar fragment and different takes out frame strategy, thus can while greatly shortening monitor video number of frames reservation useful information as much as possible.The above is the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, under the premise without departing from the principles of the invention; can also make some improvements and modifications, these improvements and modifications are also considered as protection scope of the present invention.
Claims (5)
1. extract a method for key message frame in monitor video, it is characterized in that: comprise the following steps:
S1, mark active frame/frozen frozen mass, the 1st frame of monitor video is not marked, and from the 2nd frame, the luminance difference according to present frame and former frame marks it for active frame or frozen frozen mass;
S2, time division sheet, be divided into activity time sheet (P1), sheet quiescent time (P2) and intermittently active timeslice (P3) three types video segment by monitor video;
S3, extraction frame of video, frame is taken out at the less interval of activity time sheet, and quiescent time, sheet took out frame with large-spacing, and intermittently active timeslice, when the active frame frequency of occurrences is low, take out frame with larger interval, when the more and frequency of occurrences height of active frame, take out frame with less interval.
2. extract the method for key message frame in a kind of monitor video according to claim 1, it is characterized in that: the computational methods of the mark active frame/frozen frozen mass described in step s1 are as follows:
d(t)＝∑|I
i，j(t)-I
i，j(t-1)| (1)
Wherein, I
i, jt () is the pixel brightness value of t frame position (i, j), I
i, j(t-1) be the pixel brightness value of t-1 frame position (i, j), d (t) represents the absolute value sum of t frame and its former frame correspondence position pixel intensity difference; T1 is the threshold value that user sets, and detects the size of moving object as required, can suitably change.In common application, T1 is set to 1/4 of brightness maxima, and such as 8-bit video T1 can be set to 64.When the luminance difference of t frame and former frame exceedes threshold value T1, t frame is just marked as 1 (i.e. active frame), otherwise is labeled as 0 (i.e. frozen frozen mass).
3. extract the method for key message frame in a kind of monitor video according to claim 1, it is characterized in that: activity time sheet (P1) described in step s2 is all marked as active frame and timeslice reaches the timeslice of certain length for each frame video; Described sheet quiescent time (P2) is all marked as frozen frozen mass and timeslice reaches the timeslice of certain length for each frame video; Described intermittently active timeslice (P3) is both containing active frame, also containing frozen frozen mass, but same frame of video continuous print quantity do not reach corresponding activity time sheet or quiescent time sheet the timeslice that requires of time span.
4. extract the method for key message frame in a kind of monitor video according to claim 1 and 3, it is characterized in that: described in step s2, the method for time division sheet is as follows:
Wherein, p refers to timeslice type, and N{b (t)=1} represents the quantity of continuously active frame, N{b (T)=0} represents the quantity of continuous frozen frozen mass, and T2, T3 are the threshold values that user sets, according to concrete monitoring scene and task needs, this threshold value is arranged can certain change.The basic principle of setting is that, when video primary dcreening operation, T2 establishes height, and T3 establishes height; And need to some have a mind to avoid the criminal of video monitoring do meticulous detection time, then T2 establishes low, and T3 establishes low; T2, T3 are set as the frame number of 1 second under normal circumstances.
5. extract the method for key message frame in a kind of monitor video according to claim 1, it is characterized in that: the method that step s3 extracts frame of video is as follows:
Activity time sheet (P1): take out frame by comparatively small time intervals T4, wherein T4 is user's setting threshold, advises choosing value within 0.5-2 second, is set as 1 second under normal circumstances;
Sheet quiescent time (P2): take out frame by larger time interval T5, wherein T5 is user's setting threshold, advises value more than 2 seconds, can just get 2 frames in sheet a quiescent time under normal circumstances, namely timeslice start and at the end of each frame;
Intermittently active timeslice (P3): be divided into two kinds of situations, one be in timeslice continuously active frame time length be less than 0.2 second and the ratio of active frame in timeslice lower than 20% time, take out frame by larger time interval T5, wherein T5 is user's setting threshold, advises value more than 2 seconds; Otherwise take out frame by comparatively small time intervals T6, wherein T6 is user's setting threshold, advises value more than 1 second.With activity time sheet and quiescent time sheet take out frame method unlike, be not that fixed intervals take out frame to the frame of taking out in intermittently active timeslice, but extract be greater than setting threshold after first active frame.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510062263.XA CN104618679B (en) | 2015-03-13 | 2015-03-13 | A kind of method that key message frame is extracted in monitor video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510062263.XA CN104618679B (en) | 2015-03-13 | 2015-03-13 | A kind of method that key message frame is extracted in monitor video |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104618679A true CN104618679A (en) | 2015-05-13 |
| CN104618679B CN104618679B (en) | 2018-03-27 |
Family
ID=53152938
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510062263.XA Active CN104618679B (en) | 2015-03-13 | 2015-03-13 | A kind of method that key message frame is extracted in monitor video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104618679B (en) |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106470323A (en) * | 2015-08-14 | 2017-03-01 | 杭州海康威视系统技术有限公司 | The storage method of video data and equipment |
| CN106961567A (en) * | 2016-01-12 | 2017-07-18 | 中兴通讯股份有限公司 | Animated image generation method and device |
| CN107018367A (en) * | 2017-04-11 | 2017-08-04 | 深圳市粮食集团有限公司 | A kind of method and system for implementing grain monitoring |
| CN107613237A (en) * | 2017-09-14 | 2018-01-19 | 国网重庆市电力公司电力科学研究院 | A kind of abstracting method of video sound mixing key frame |
| CN107666610A (en) * | 2017-09-25 | 2018-02-06 | 北京小鱼易连科技有限公司 | Desktop video content analysis method |
| CN108377399A (en) * | 2018-03-07 | 2018-08-07 | 广州图普网络科技有限公司 | Live video stream code-transferring method, device and computer readable storage medium |
| CN110582016A (en) * | 2019-09-06 | 2019-12-17 | 北京达佳互联信息技术有限公司 | video information display method, device, server and storage medium |
| CN111385601A (en) * | 2018-12-29 | 2020-07-07 | 华为技术有限公司 | Video auditing method and system |
| CN111835955A (en) * | 2019-04-15 | 2020-10-27 | 杭州海康威视数字技术股份有限公司 | Data acquisition method and device |
| CN112073721A (en) * | 2020-10-10 | 2020-12-11 | 绍兴文理学院 | High-efficiency non-I frame image coding and decoding method |
| CN112866646A (en) * | 2021-01-13 | 2021-05-28 | 杭州雅观科技有限公司 | Smart community data acquisition method and system based on Bluetooth Mesh |
| CN112989113A (en) * | 2021-04-21 | 2021-06-18 | 北京金和网络股份有限公司 | Picture screening method, device and equipment |
| CN113365018A (en) * | 2020-03-06 | 2021-09-07 | 杭州海康威视数字技术股份有限公司 | Period adjusting method, device and equipment |
| CN113556582A (en) * | 2021-07-30 | 2021-10-26 | 海宁奕斯伟集成电路设计有限公司 | Video data processing method, device, equipment and storage medium |
| CN113573061A (en) * | 2020-04-29 | 2021-10-29 | 安徽华米健康科技有限公司 | Video frame extraction method, device and equipment |
| CN113612952A (en) * | 2021-08-11 | 2021-11-05 | 四川亨通网智科技有限公司 | Method for adjusting dynamic allocation of AI server resources based on frequency conversion example |
| CN115065798A (en) * | 2022-08-18 | 2022-09-16 | 广州智算信息技术有限公司 | Big data-based video analysis monitoring system |
| CN116805433A (en) * | 2023-06-27 | 2023-09-26 | 北京奥康达体育科技有限公司 | Human motion trail data analysis system |
| CN117119253A (en) * | 2023-06-28 | 2023-11-24 | 三峡科技有限责任公司 | High-quality video frame extraction method for target object |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030095785A1 (en) * | 2001-11-16 | 2003-05-22 | Keisuke Izumi | Digital image processing apparatus, digital image processing method, digital image processing program product, and digital image printing system |
| CN101026752A (en) * | 2007-03-15 | 2007-08-29 | 杭州华为三康技术有限公司 | Method and device for processing frame transmission rate, and computer software product |
| JP2009294983A (en) * | 2008-06-06 | 2009-12-17 | Sony Corp | Tracking point detection device and method, program, and recording medium |
| CN101827264A (en) * | 2009-03-06 | 2010-09-08 | 刘永 | Hierarchical self-adaptive video frame sampling method |
| CN102930513A (en) * | 2012-09-25 | 2013-02-13 | 北京航空航天大学 | Method for integrating virtual light and real light of video scene |
-
2015
- 2015-03-13 CN CN201510062263.XA patent/CN104618679B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030095785A1 (en) * | 2001-11-16 | 2003-05-22 | Keisuke Izumi | Digital image processing apparatus, digital image processing method, digital image processing program product, and digital image printing system |
| CN101026752A (en) * | 2007-03-15 | 2007-08-29 | 杭州华为三康技术有限公司 | Method and device for processing frame transmission rate, and computer software product |
| JP2009294983A (en) * | 2008-06-06 | 2009-12-17 | Sony Corp | Tracking point detection device and method, program, and recording medium |
| CN101827264A (en) * | 2009-03-06 | 2010-09-08 | 刘永 | Hierarchical self-adaptive video frame sampling method |
| CN102930513A (en) * | 2012-09-25 | 2013-02-13 | 北京航空航天大学 | Method for integrating virtual light and real light of video scene |
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106470323B (en) * | 2015-08-14 | 2019-08-16 | 杭州海康威视系统技术有限公司 | The storage method and equipment of video data |
| CN106470323A (en) * | 2015-08-14 | 2017-03-01 | 杭州海康威视系统技术有限公司 | The storage method of video data and equipment |
| CN106961567A (en) * | 2016-01-12 | 2017-07-18 | 中兴通讯股份有限公司 | Animated image generation method and device |
| CN107018367A (en) * | 2017-04-11 | 2017-08-04 | 深圳市粮食集团有限公司 | A kind of method and system for implementing grain monitoring |
| CN107613237B (en) * | 2017-09-14 | 2020-03-06 | 国网重庆市电力公司电力科学研究院 | Extraction method of video dynamic and static mixed key frames |
| CN107613237A (en) * | 2017-09-14 | 2018-01-19 | 国网重庆市电力公司电力科学研究院 | A kind of abstracting method of video sound mixing key frame |
| CN107666610A (en) * | 2017-09-25 | 2018-02-06 | 北京小鱼易连科技有限公司 | Desktop video content analysis method |
| CN108377399A (en) * | 2018-03-07 | 2018-08-07 | 广州图普网络科技有限公司 | Live video stream code-transferring method, device and computer readable storage medium |
| CN108377399B (en) * | 2018-03-07 | 2019-03-19 | 广州图普网络科技有限公司 | Live video stream code-transferring method, device and computer readable storage medium |
| CN111385601A (en) * | 2018-12-29 | 2020-07-07 | 华为技术有限公司 | Video auditing method and system |
| CN111835955B (en) * | 2019-04-15 | 2021-12-21 | 杭州海康威视数字技术股份有限公司 | Data acquisition method and device |
| CN111835955A (en) * | 2019-04-15 | 2020-10-27 | 杭州海康威视数字技术股份有限公司 | Data acquisition method and device |
| CN110582016A (en) * | 2019-09-06 | 2019-12-17 | 北京达佳互联信息技术有限公司 | video information display method, device, server and storage medium |
| CN113365018B (en) * | 2020-03-06 | 2022-07-01 | 杭州海康威视数字技术股份有限公司 | Period adjusting method, device and equipment |
| CN113365018A (en) * | 2020-03-06 | 2021-09-07 | 杭州海康威视数字技术股份有限公司 | Period adjusting method, device and equipment |
| CN113573061A (en) * | 2020-04-29 | 2021-10-29 | 安徽华米健康科技有限公司 | Video frame extraction method, device and equipment |
| CN112073721A (en) * | 2020-10-10 | 2020-12-11 | 绍兴文理学院 | High-efficiency non-I frame image coding and decoding method |
| CN112073721B (en) * | 2020-10-10 | 2024-02-27 | 绍兴文理学院 | Efficient non-I frame image coding and decoding method |
| CN112866646A (en) * | 2021-01-13 | 2021-05-28 | 杭州雅观科技有限公司 | Smart community data acquisition method and system based on Bluetooth Mesh |
| CN112989113B (en) * | 2021-04-21 | 2021-08-31 | 北京金和网络股份有限公司 | Picture screening method, device and equipment |
| CN112989113A (en) * | 2021-04-21 | 2021-06-18 | 北京金和网络股份有限公司 | Picture screening method, device and equipment |
| CN113556582A (en) * | 2021-07-30 | 2021-10-26 | 海宁奕斯伟集成电路设计有限公司 | Video data processing method, device, equipment and storage medium |
| CN113612952A (en) * | 2021-08-11 | 2021-11-05 | 四川亨通网智科技有限公司 | Method for adjusting dynamic allocation of AI server resources based on frequency conversion example |
| CN115065798A (en) * | 2022-08-18 | 2022-09-16 | 广州智算信息技术有限公司 | Big data-based video analysis monitoring system |
| CN115065798B (en) * | 2022-08-18 | 2022-11-22 | 广州智算信息技术有限公司 | Big data-based video analysis monitoring system |
| CN116805433A (en) * | 2023-06-27 | 2023-09-26 | 北京奥康达体育科技有限公司 | Human motion trail data analysis system |
| CN116805433B (en) * | 2023-06-27 | 2024-02-13 | 北京奥康达体育科技有限公司 | Human motion trail data analysis system |
| CN117119253A (en) * | 2023-06-28 | 2023-11-24 | 三峡科技有限责任公司 | High-quality video frame extraction method for target object |
| CN117119253B (en) * | 2023-06-28 | 2024-05-07 | 三峡科技有限责任公司 | High-quality video frame extraction method for target object |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104618679B (en) | 2018-03-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104618679A (en) | Method for extracting key information frame from monitoring video | |
| CN104639860B (en) | The storage method and device of a kind of monitoring video | |
| CN108875561A (en) | A kind of hidden danger repetition method of discrimination of transmission line of electricity monitoring hidden danger early warning image | |
| CN103326885B (en) | A kind of processing method and processing device of warning message | |
| CN103902476B (en) | A kind of webpage back door detection method and system based on non-credit | |
| CN108629322A (en) | Monitoring method, device and the computer readable storage medium of a suspect | |
| CN103166806B (en) | Plug-in program detection method and the system of third party's application | |
| CN105227927A (en) | A kind of monitor video date storage method and device | |
| CN105472412B (en) | A kind of big data processing method for distinguishing smart television state | |
| CN110738178A (en) | Garden construction safety detection method and device, computer equipment and storage medium | |
| CN104850841A (en) | Elder abnormal behavior monitoring method with combination of RFID and video identification | |
| CN104378361A (en) | Network intrusion detection method and system | |
| Ullah et al. | Gaussian mixtures for anomaly detection in crowded scenes | |
| CN101141292B (en) | Method of implementing alarm automatic refreshment in network management system | |
| CN107547941A (en) | Storage method, the device and system of media data | |
| CN111881946A (en) | Safety monitoring method and device, storage medium, electronic equipment and air conditioner | |
| CN103761879B (en) | A kind of counterfeit vehicle registration plate identification method and system | |
| US11830286B2 (en) | Data processing apparatus, data processing method, and non-transitory storage medium | |
| CN114173094A (en) | Video monitoring method and device, computer equipment and storage medium | |
| CN103824074A (en) | Crowd density estimation method based on background subtraction and texture features and system | |
| CN107146372A (en) | A kind of method and system by video identification production line working condition | |
| CN106354846A (en) | Intelligent news manuscript selection method and system based on big data | |
| CN111080139A (en) | Hypergraph-based public security event scenario representation method | |
| CN106227823A (en) | A kind of webpage update detection method, info web capture and rendering method | |
| CN114390260A (en) | Hazardous area and important place monitoring platform applied to smart city |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |