CN103929685A - Video abstract generating and indexing method - Google Patents
Video abstract generating and indexing method Download PDFInfo
- Publication number
- CN103929685A CN103929685A CN201410151449.8A CN201410151449A CN103929685A CN 103929685 A CN103929685 A CN 103929685A CN 201410151449 A CN201410151449 A CN 201410151449A CN 103929685 A CN103929685 A CN 103929685A
- Authority
- CN
- China
- Prior art keywords
- video
- target
- targets
- moving
- background
- 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
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000000605 extraction Methods 0.000 claims abstract description 9
- 230000000694 effects Effects 0.000 claims description 9
- 230000015572 biosynthetic process Effects 0.000 claims description 7
- 238000003786 synthesis reaction Methods 0.000 claims description 7
- 230000007547 defect Effects 0.000 claims description 5
- 229940037201 oris Drugs 0.000 claims description 3
- 230000000877 morphologic effect Effects 0.000 claims description 2
- 230000002194 synthesizing effect Effects 0.000 claims description 2
- 239000011800 void material Substances 0.000 claims 2
- 230000000052 comparative effect Effects 0.000 abstract 1
- 230000011218 segmentation Effects 0.000 abstract 1
- 238000012544 monitoring process Methods 0.000 description 8
- 238000012937 correction Methods 0.000 description 5
- 239000000203 mixture Substances 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000011840 criminal investigation Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005315 distribution function Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to a video abstract generating and indexing method. The video abstract generating and indexing method comprises the following steps of (1) conducting background modeling on a target image in an original video, and achieving background extraction; (2) conducting comparative segmentation on a current image and a background model, and determining movement targets according to a comparison result; (3) matching space information distributed between frames with characteristics of the targets separated from each image, achieving target tracking and recording the movement trails of the targets; (4) correcting the positions of the targets in the images in a sequence; (5) allowing the background to be overlaid with the movement targets to form a short video, and recording abstract video data to form an index data file. The video abstract generating and indexing method has the advantages that after the movement targets are tracked, the tracking success rate is guaranteed, the quality of a generated video abstract is improved greatly, and the video indexing function can meet the requirements that users view the video rapidly and view the original video conveniently so as to view actual situations completely.
Description
Technical Field
The invention relates to the technical field of video monitoring, in particular to a video abstract generating and indexing method.
Background
With the rapid development of multimedia technology and video acquisition technology, video monitoring, image compression coding and streaming media technology gradually evolves, and the application of video monitoring technology in daily life is more and more extensive, so that video monitoring is not limited to safety precaution, but becomes an effective supervision means for all industries, and the flexibility of the application field of the video monitoring technology far exceeds the scope defined by the traditional security monitoring. However, video monitoring videos have the characteristics of large data storage amount, long storage time and the like, clues are searched through the videos to obtain evidences, a large amount of manpower, material resources and time are consumed in the traditional method, and the efficiency is extremely low, so that the best opportunity is missed. Therefore, in the video monitoring system, the original video is concentrated, so that the retrieval object can be quickly browsed and locked, and various requirements and applications of public security, network supervision and criminal investigation can be met. Video summarization plays an important role in video analysis and content-based video retrieval, which generates a short video that contains all the important activities in the original video. The video compresses the whole original video into a short summary of events by playing multiple events simultaneously, even if they occur at different times in the original video.
The purpose of the video abstract is to facilitate a user to quickly view a video, and the quality of the generated video abstract directly influences the experience effect of the user. At present, the problems of incomplete targets, ghosts and the like commonly exist in video abstractions. Moreover, the summarized video disturbs the sequential logic of the target in the original video, and a user needs to view the target by means of the original video if the user needs to know the real situation of a certain target, so that the problem that how to directly jump from the summarized video to the original video to view the situation of the target is also solved.
Disclosure of Invention
The invention aims to provide a video abstract generating and indexing method to overcome the defects in the prior art.
The purpose of the invention is realized by the following technical scheme:
a video abstract generating and indexing method comprises the following steps:
1) background modeling: performing background modeling on a target frame image in an original video to realize background extraction and separating a background from the original video;
2) extracting a moving target: comparing and segmenting the current image and the background model, and determining a moving target according to a comparison result;
3) tracking a moving target: matching spatial information distributed between frames with the characteristics of a target segmented from each frame of image to realize target tracking and recording the motion track of the target;
4) and (3) correcting the position of the moving target: correcting the tracked target set, mainly correcting the position of a target in the sequence in the image; and
5) abstract synthesis and video index establishment: the method comprises the steps of superposing moving objects on a background, synchronously playing activities which do not occur simultaneously in an original video under the condition that the video summary has no shielding or has small shielding, generating a summary video which is relatively compact in time and space and contains necessary activities in the original video, and recording summary video data when synthesizing the videos to form an index data file.
Further, the background modeling in the step 1) adopts a color background model method.
Preferably, the color background model specifically adopts a mixed gaussian background algorithm.
Further, if the moving target extracted in the step 2) has the defect of cavity and noise interference, performing morphological opening and closing operation to eliminate the cavity and noise; if the moving target extracted in the step 2) has the defect that the same target is divided into two or more targets, calculating the mutual space distance between the targets for all the targets extracted in each frame, and identifying the targets with the distance smaller than the threshold value lambda as the same target.
Further, the step 3) specifically comprises the following steps:
a) the tracking module performs matching tracking between moving targets of adjacent frames by using the spatial distribution information and the color characteristics, the successfully matched moving targets are regarded as the same target, and the moving track is recorded; unsuccessful matching is considered as a new moving object.
b) Storing the tracked results in a collectionIn (1),the target is expressed as follows:
wherein,representing objectsThe sequence occurring in the video.
Further, the method for matching and tracking comprises the following two steps:
the first method comprises the following steps: matching the target segmented from a new frame with the targets in the set omega, and defining the following functions:
time difference function:
wherein,a newly extracted object is represented that is,representation collectionOne object of (1).To representThe time stamp of (a) is stored,to representThe time stamp of (c).Is a defined time difference threshold.
Distance difference function:
wherein,a newly extracted object is represented that is,representation collectionOne object of (1).To representAnddistance in space.Is a defined distance difference threshold.
A comparison function:
if the function is comparedIs 1, then calculateAndif the histogram distance threshold is met, the matching is successful, and the color histogram distance of the color histogram is calculatedIs added toIn the sequence of (a). If the match is unsuccessful orIs 0, thenIs a new target to beAdding to collectionsIn (1).
And the second method comprises the following steps: according to the first method, the target beta is first combined withLast frame of target sequenceComparing, if the matching is not successful, then the comparison is performedUntil the previous frame is comparedThe first M frames.
Further, the step 4) of correcting the position of the moving target specifically includes the following steps:
firstly, after the video is completely processed, countingEach of the objects inWidth, height and ordering of objects in the sequence of (a).
After sortingThe widths of (a) are expressed as follows:
after sortingIs expressed as follows:
secondly, respectively calculating the average value of the sequences to obtain the targetWidth of (2)And heightAccording toAndeach target position in the target sequence is corrected.
Further, during the summary synthesis in step 5), the encoding, position and first occurrence timestamp of the moving object participating in the combination in each frame of the video summary are recorded, and these values are maintained in the index file.
The invention has the beneficial effects that: after the moving target is tracked, the tracking success rate is ensured, the quality of the generated video abstract is greatly improved, and the video indexing function can meet the requirement that a user quickly checks the video, conveniently checks the original video and completely watches the actual situation.
Drawings
The invention is explained in further detail below with reference to the drawing.
Fig. 1 is a schematic flowchart of a video summary generation and indexing method according to an embodiment of the present invention;
FIG. 2 illustrates the position of a sequence of moving objects in an image before modification according to an embodiment of the present invention;
fig. 3 shows the position of a modified sequence of moving objects in an image according to an embodiment of the present invention.
Detailed Description
As shown in fig. 1, the steps of the embodiment of the present invention include background modeling, moving object extraction, moving object tracking, moving object correction, summary synthesis, and video indexing. The method comprises the following specific steps:
1. background modeling
The background modeling module may use various image background modeling algorithms, including both color background models and texture background models. The idea of the color background model is to model the color value (grayscale or color) of each pixel in the image. If the pixel color value at the current image coordinate (x, y) is different from the pixel color value at (x, y) in the background model, the current pixel is considered as foreground, otherwise, the current pixel is considered as background.
The background modeling module of the present embodiment uses a Gaussian Mixture background algorithm in a color background Model, and a Gaussian Mixture background Model (Gaussian Mixture Model) is developed based on a single Gaussian Model, and a density distribution function of an arbitrary shape is smoothly approximated by a weighted average of a plurality of Gaussian probability density functions. The Gaussian mixture model assumes that the Gaussian distribution for describing the color of each pixel point is K, and generally takes 3-5. The number of K values in this embodiment is 3.
2. Moving object extraction
After the background model is established, the current image is compared with the background model in a certain way, and the moving target to be detected is determined according to the comparison result. In general, the obtained foreground includes much noise, and in order to eliminate the noise, the present embodiment performs an opening operation and a closing operation on the extracted moving target image, and then discards a relatively small contour.
After the targets are extracted, the total number of pixels included in each target is counted, and if the total number of pixels of a certain target is less than 400 pixels, the target is regarded as interference elimination and is not processed.
In order to solve the problem of dividing the same target into two or more targets, the mutual space distance between all targets in the current frame is calculated, and the distance is smaller than the distance by taking a pixel as a unitAre identified as the same object. In this embodiment, Λ is 15 pixels.
3. Moving object tracking module
For a certain moving target of the current frame, because the inter-frame time interval is short, the size of the space occupied by the moving target and the change of the position of the space occupied by the moving target are small, and the embodiment performs matching tracking between the moving targets of adjacent frames by using the space distribution information and the color characteristics.
The tracking module performs matching tracking between moving objects of adjacent frames by using the spatial distribution information and the color features. And if the matching is successful, the same object is considered, the motion trail is recorded, and if the matching is unsuccessful, a new motion object is considered.
The results of the tracing are stored in a collectionIn (1),the target is expressed as follows:
wherein,representing objectsThe sequence occurring in the video.
If the target extraction effect of a certain frame of the extraction module is not good, the tracking failure can be caused. In order to improve the success rate of tracking, the following two methods are adopted:
1) the targets divided from the new frame are not matched with the targets of the previous frame, but are aggregatedThe following functions are defined:
time difference function:
wherein,a newly extracted object is represented that is,representation collectionOne object of (1).To representThe time stamp of (a) is stored,to representThe time stamp of (c).Is a defined time difference threshold.
Distance difference function:
wherein,a newly extracted object is represented that is,representation collectionOne object of (1).To representAnddistance in space.Is a defined distance difference threshold.
A comparison function:
if the function is comparedIs 1, then calculateAndif the histogram distance threshold is met, the matching is successful, and the color histogram distance of the color histogram is calculatedIs added toIn the sequence of (a). If the match is unsuccessful orIs 0, thenIs a new target to beAdding to collectionsIn (1).
2) In the above method, only the objects are combinedThe latest frame of the target sequence is compared ifThe last frame is not well extracted, and the tracking failure can occur. First, target beta andlast frame of target sequenceComparing, if the matching is not successful, then the comparison is performedUntil the previous frame is comparedThe first M frames.
In this embodiment, the frame number of the target appearing in the video stream sequence is taken as its time stamp, and the first frame number is 0, and sequentially increases. Time difference function of the present embodimentThe value is 15, which represents the target to be matchedAnd collectionsMiddle targetShould be within 15 frames.
In the present embodiment, the target to be matched is calculatedAnd collectionsMiddle targetThe distance between the two objects is the pixel value between the closest point of the distance between the two objects, and the distance difference functionA value of 20 represents the target to be matchedAnd collectionsMiddle targetShould be within 20 pixels.
In the tracking module of the embodiment, the value of m is 10, which represents the target to be matchedCan be combined withMiddle targetThe last 10 targets of the sequence are compared in reverse order of their time of occurrence.
The embodiment statistically obtains the target to be matchedAnd collectionsInThe Bhattacharyya distance of the two histograms is calculated to describe the similarity of the two histograms. If the Bhattacharyya distance is less than 0.6, the result shows thatAnd collectionsInThe matching is successful, willIs added toIn (c) sequence (c). If it is notAndall the targets in (1) can not be matched, thenAn object code is toAdd into collectionsIn (1).
4. Moving object position correction module
After the video is completely processed, the moving target position correction module carries out statisticsEach of the objects inWidth, height of the targets in the sequence of (1), versus targets in [ omega ]Correcting the target position of the sequence, toThe width and height in the sequence are sorted from large to small, and after sorting, the sequence is divided into a plurality of sectionsThe widths of (a) are expressed as follows:
after sortingIs expressed as follows:
and calculating the average value of the N front widths and heights after sorting to obtain the average width and the average height. Where N is taken to be20% of the total sequence. Root of correct timeAnd according to the principle of target center alignment, the target width is symmetrically modified from left to right, and the target height is symmetrically modified from top to bottom.
The position of the target sequence in the image before the correction is shown in fig. 2, and the position of the target sequence after the correction is shown in fig. 3. After the position of the moving target is corrected, the problem of incomplete target extraction in the extraction process can be solved, and the quality of the generated video abstract is improved.
5. Summary synthesis and video indexing
The module mainly completes the synthesis of the tracked moving object and the video background, and synchronously plays the activities which do not occur simultaneously in the original video under the condition of no occlusion (or small occlusion) in the video abstract, thereby generating an abstract video which is relatively compact in time and space and contains the necessary activities in the original video.
For each frame of image of the video summary, selecting which moving objects appear simultaneously is the key to the composition. The present embodiment is determined by calculating an energy loss function for each moving object. The function consists of a moving target time difference loss function and a moving target collision loss function, and moving targets with energy loss function values meeting conditions are selected and combined.
Before generating the video abstract of each frame, dividing the moving objects into three types of sets: merged complete (S1), merging in progress (S2), to-be-merged (S3). And sequentially calculating energy loss functions between the sum set S2 from S3 according to the sequence of the occurrence time, and merging the energy loss functions in the same frame of video when the energy loss functions meet the loss threshold.
When merging, a background image needs to be provided, and the background at the earliest time of the occurrence time of the moving object in the frame is selected as the background image.
When merging the moving objects, recording the codes, positions and time stamps appearing for the first time of the moving objects participating in merging in each frame, and storing the values in an index file.
When a user clicks a video, judging whether the position of a mouse is within an envelope range of a moving target, and if the position of the mouse is within a certain target range, inquiring an index file to obtain the time of the target appearing in the original video.
The present invention is not limited to the above-mentioned preferred embodiments, and any other products in various forms can be obtained by anyone in the light of the present invention, but any changes in the shape or structure thereof, which have the same or similar technical solutions as those of the present application, fall within the protection scope of the present invention.
Claims (8)
1. A video abstract generating and indexing method is characterized by comprising the following steps:
1) background modeling: performing background modeling on a target frame image in an original video to realize background extraction and separating a background from the original video;
2) extracting a moving target: comparing and segmenting the current image and the background model, and determining a moving target according to a comparison result;
3) tracking a moving target: matching spatial information distributed between frames with the characteristics of a target segmented from each frame of image to realize target tracking and recording the motion track of the target;
4) and (3) correcting the position of the moving target: correcting the tracked target set, mainly correcting the position of a target in the sequence in the image; and
5) abstract synthesis and video index establishment: the method comprises the steps of superposing moving objects on a background, synchronously playing activities which do not occur simultaneously in an original video under the condition that the video summary has no shielding or has small shielding, generating a summary video which is relatively compact in time and space and contains necessary activities in the original video, and recording summary video data when synthesizing the videos to form an index data file.
2. The method for generating and indexing video summary according to claim 1, wherein in step 1), the background modeling adopts a color background model.
3. The method for video summary generation and indexing as claimed in claim 2, wherein: the color background model specifically adopts a mixed Gaussian background algorithm.
4. The video abstract generating and indexing method according to claim 1, wherein in the step 2), if the extracted moving object has the defect of void and noise interference, the morphological open and close operation is adopted for processing to eliminate the void and the noise; if the moving target extracted in the step 2) has the defect that the same target is divided into two or more targets, calculating the mutual space distance between the targets for all the targets extracted in each frame, and identifying the targets with the distance smaller than the threshold value lambda as the same target.
5. The method for generating and indexing a video summary according to claim 1, wherein the step 3) comprises the following steps:
a) the tracking module performs matching tracking between moving targets of adjacent frames by using the spatial distribution information and the color characteristics, the successfully matched moving targets are regarded as the same target, and the moving track is recorded; the matching is unsuccessful and is regarded as a new moving object; and
b) storing the tracked results in a collectionIn (1),the target is expressed as follows:
wherein,representing objectsThe sequence occurring in the video.
6. The method for video summary generation and indexing as claimed in claim 5, wherein the matching tracking method includes the following two methods:
the first method comprises the following steps: matching the target segmented from a new frame with the targets in the set omega, and defining the following functions:
time difference function:
wherein,a newly extracted object is represented that is,representation collectionOne of the targets;to representThe time stamp of (a) is stored,to representA timestamp of (d);is a defined time difference threshold;
distance difference function:
wherein,a newly extracted object is represented that is,representation collectionIs the object of (1) a method for,to representAnda distance in space;is a defined distance difference threshold;
a comparison function:
if the function is comparedIs 1, then calculateAndif the histogram distance threshold is met, the matching is successful, and the color histogram distance of the color histogram is calculatedIs added toIn the sequence of (a); if the match is unsuccessful orIs 0, thenIs a new oneTarget ofAdding to collectionsPerforming the following steps;
and the second method comprises the following steps: according to the first method, the target beta is first combined withLast frame of target sequenceComparing, if the matching is not successful, then the comparison is performedUntil the previous frame is comparedThe first M frames.
7. The method for generating and indexing a video summary according to claim 1, wherein the step 4) of correcting the position of the moving object specifically comprises the following steps:
firstly, after the video is completely processed, countingEach of the objects inWidth, height and ordering of the targets in the sequence of (a);
after sortingThe widths of (a) are expressed as follows:
after sortingIs expressed as follows:
secondly, respectively calculating the average value of the sequences to obtain the targetWidth of (2)And heightAccording toAnd correcting each target position in the target sequence.
8. The method for video summary generation and indexing as claimed in claim 1, wherein: in step 5), during summary synthesis, the encoding, position and first-appearing timestamp of the moving object participating in the combination in each frame of the video summary need to be recorded, and the values are kept in the index file.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410151449.8A CN103929685B (en) | 2014-04-15 | 2014-04-15 | A kind of video frequency abstract generation and indexing means |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410151449.8A CN103929685B (en) | 2014-04-15 | 2014-04-15 | A kind of video frequency abstract generation and indexing means |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103929685A true CN103929685A (en) | 2014-07-16 |
| CN103929685B CN103929685B (en) | 2017-11-07 |
Family
ID=51147740
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410151449.8A Active CN103929685B (en) | 2014-04-15 | 2014-04-15 | A kind of video frequency abstract generation and indexing means |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103929685B (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104268563A (en) * | 2014-09-15 | 2015-01-07 | 合肥工业大学 | Video abstraction method based on abnormal behavior detection |
| CN104717573A (en) * | 2015-03-05 | 2015-06-17 | 广州市维安电子技术有限公司 | Video abstract generation method |
| CN104935888A (en) * | 2015-06-11 | 2015-09-23 | 惠州Tcl移动通信有限公司 | Video monitoring method capable of marking object and video monitoring system thereof |
| CN105007464A (en) * | 2015-07-20 | 2015-10-28 | 江西洪都航空工业集团有限责任公司 | Method for concentrating video |
| CN105469425A (en) * | 2015-11-24 | 2016-04-06 | 上海君是信息科技有限公司 | Video condensation method |
| CN106714007A (en) * | 2016-12-15 | 2017-05-24 | 重庆凯泽科技股份有限公司 | Video abstract method and apparatus |
| CN107493520A (en) * | 2016-06-13 | 2017-12-19 | 合肥君正科技有限公司 | A kind of video abstraction generating method and device |
| CN107529098A (en) * | 2014-09-04 | 2017-12-29 | 英特尔公司 | Real-time video is made a summary |
| CN107920213A (en) * | 2017-11-20 | 2018-04-17 | 深圳市堇茹互动娱乐有限公司 | Image synthesizing method, terminal and computer-readable recording medium |
| CN109511019A (en) * | 2017-09-14 | 2019-03-22 | 中兴通讯股份有限公司 | A kind of video summarization method, terminal and computer readable storage medium |
| CN110532916A (en) * | 2019-08-20 | 2019-12-03 | 北京地平线机器人技术研发有限公司 | A kind of motion profile determines method and device |
| WO2020037881A1 (en) * | 2018-08-21 | 2020-02-27 | 北京字节跳动网络技术有限公司 | Motion trajectory drawing method and apparatus, and device and storage medium |
| CN111539974A (en) * | 2020-04-07 | 2020-08-14 | 北京明略软件系统有限公司 | Method and device for determining track, computer storage medium and terminal |
| CN111563489A (en) * | 2020-07-14 | 2020-08-21 | 浙江大华技术股份有限公司 | Target tracking method and device and computer storage medium |
| CN111739128A (en) * | 2020-07-29 | 2020-10-02 | 广州筷子信息科技有限公司 | Target video generation method and system |
| CN111918146A (en) * | 2020-07-28 | 2020-11-10 | 广州筷子信息科技有限公司 | Video synthesis method and system |
| CN113469200A (en) * | 2020-03-30 | 2021-10-01 | 阿里巴巴集团控股有限公司 | Data processing method and system, storage medium and computing device |
| CN113873200A (en) * | 2021-09-26 | 2021-12-31 | 珠海研果科技有限公司 | Image identification method and system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004047109A1 (en) * | 2002-11-19 | 2004-06-03 | Koninklijke Philips Electronics N.V. | Video abstracting |
| CN202003350U (en) * | 2010-12-14 | 2011-10-05 | 广东鑫程电子科技有限公司 | Video summary system |
| CN102930061A (en) * | 2012-11-28 | 2013-02-13 | 安徽水天信息科技有限公司 | Video abstraction method and system based on moving target detection |
| CN103092929A (en) * | 2012-12-30 | 2013-05-08 | 信帧电子技术（北京）有限公司 | Method and device for generation of video abstract |
| CN103092963A (en) * | 2013-01-21 | 2013-05-08 | 信帧电子技术（北京）有限公司 | Video abstract generating method and device |
| EP2640084A2 (en) * | 2012-03-14 | 2013-09-18 | Kabushiki Kaisha Toshiba | Video distribution system, apparatus, method and medium for distributing time shifted content starting from a selected chapter |
| CN103473333A (en) * | 2013-09-18 | 2013-12-25 | 北京声迅电子股份有限公司 | Method and device for extracting video abstract from ATM (Automatic Teller Machine) scene |
-
2014
- 2014-04-15 CN CN201410151449.8A patent/CN103929685B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004047109A1 (en) * | 2002-11-19 | 2004-06-03 | Koninklijke Philips Electronics N.V. | Video abstracting |
| CN202003350U (en) * | 2010-12-14 | 2011-10-05 | 广东鑫程电子科技有限公司 | Video summary system |
| EP2640084A2 (en) * | 2012-03-14 | 2013-09-18 | Kabushiki Kaisha Toshiba | Video distribution system, apparatus, method and medium for distributing time shifted content starting from a selected chapter |
| CN102930061A (en) * | 2012-11-28 | 2013-02-13 | 安徽水天信息科技有限公司 | Video abstraction method and system based on moving target detection |
| CN103092929A (en) * | 2012-12-30 | 2013-05-08 | 信帧电子技术（北京）有限公司 | Method and device for generation of video abstract |
| CN103092963A (en) * | 2013-01-21 | 2013-05-08 | 信帧电子技术（北京）有限公司 | Video abstract generating method and device |
| CN103473333A (en) * | 2013-09-18 | 2013-12-25 | 北京声迅电子股份有限公司 | Method and device for extracting video abstract from ATM (Automatic Teller Machine) scene |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107529098A (en) * | 2014-09-04 | 2017-12-29 | 英特尔公司 | Real-time video is made a summary |
| US10755105B2 (en) | 2014-09-04 | 2020-08-25 | Intel Corporation | Real time video summarization |
| CN104268563A (en) * | 2014-09-15 | 2015-01-07 | 合肥工业大学 | Video abstraction method based on abnormal behavior detection |
| CN104268563B (en) * | 2014-09-15 | 2017-05-17 | 合肥工业大学 | Video abstraction method based on abnormal behavior detection |
| CN104717573A (en) * | 2015-03-05 | 2015-06-17 | 广州市维安电子技术有限公司 | Video abstract generation method |
| CN104717573B (en) * | 2015-03-05 | 2018-04-13 | 广州市维安电子技术有限公司 | A kind of generation method of video frequency abstract |
| CN104935888A (en) * | 2015-06-11 | 2015-09-23 | 惠州Tcl移动通信有限公司 | Video monitoring method capable of marking object and video monitoring system thereof |
| CN105007464A (en) * | 2015-07-20 | 2015-10-28 | 江西洪都航空工业集团有限责任公司 | Method for concentrating video |
| CN105469425A (en) * | 2015-11-24 | 2016-04-06 | 上海君是信息科技有限公司 | Video condensation method |
| CN107493520A (en) * | 2016-06-13 | 2017-12-19 | 合肥君正科技有限公司 | A kind of video abstraction generating method and device |
| CN106714007A (en) * | 2016-12-15 | 2017-05-24 | 重庆凯泽科技股份有限公司 | Video abstract method and apparatus |
| CN109511019A (en) * | 2017-09-14 | 2019-03-22 | 中兴通讯股份有限公司 | A kind of video summarization method, terminal and computer readable storage medium |
| CN107920213A (en) * | 2017-11-20 | 2018-04-17 | 深圳市堇茹互动娱乐有限公司 | Image synthesizing method, terminal and computer-readable recording medium |
| WO2020037881A1 (en) * | 2018-08-21 | 2020-02-27 | 北京字节跳动网络技术有限公司 | Motion trajectory drawing method and apparatus, and device and storage medium |
| US11514625B2 (en) | 2018-08-21 | 2022-11-29 | Beijing Bytedance Network Technology Co., Ltd. | Motion trajectory drawing method and apparatus, and device and storage medium |
| CN110532916A (en) * | 2019-08-20 | 2019-12-03 | 北京地平线机器人技术研发有限公司 | A kind of motion profile determines method and device |
| CN113469200A (en) * | 2020-03-30 | 2021-10-01 | 阿里巴巴集团控股有限公司 | Data processing method and system, storage medium and computing device |
| CN111539974B (en) * | 2020-04-07 | 2022-11-11 | 北京明略软件系统有限公司 | Method and device for determining track, computer storage medium and terminal |
| CN111539974A (en) * | 2020-04-07 | 2020-08-14 | 北京明略软件系统有限公司 | Method and device for determining track, computer storage medium and terminal |
| CN111563489A (en) * | 2020-07-14 | 2020-08-21 | 浙江大华技术股份有限公司 | Target tracking method and device and computer storage medium |
| CN111918146A (en) * | 2020-07-28 | 2020-11-10 | 广州筷子信息科技有限公司 | Video synthesis method and system |
| CN111739128A (en) * | 2020-07-29 | 2020-10-02 | 广州筷子信息科技有限公司 | Target video generation method and system |
| CN113873200A (en) * | 2021-09-26 | 2021-12-31 | 珠海研果科技有限公司 | Image identification method and system |
| CN113873200B (en) * | 2021-09-26 | 2024-02-02 | 珠海研果科技有限公司 | Image identification method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103929685B (en) | 2017-11-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103929685B (en) | A kind of video frequency abstract generation and indexing means | |
| Liu et al. | Key frame extraction from MPEG video stream | |
| EP2224357A1 (en) | Video segmentation | |
| JP5420199B2 (en) | Video analysis device, video analysis method, digest automatic creation system and highlight automatic extraction system | |
| Liu et al. | Effective and scalable video copy detection | |
| CN105049875B (en) | A kind of accurate extraction method of key frame based on composite character and abrupt climatic change | |
| CN104837031B (en) | A kind of method of high-speed adaptive extraction key frame of video | |
| CN1337828A (en) | Method for fetching high-grade characterstics from low-grade characteristics of multi-media content | |
| Lai et al. | Video summarization of surveillance cameras | |
| Omidyeganeh et al. | Video keyframe analysis using a segment-based statistical metric in a visually sensitive parametric space | |
| Asim et al. | A key frame based video summarization using color features | |
| CN116363554A (en) | Method, system, medium, equipment and terminal for extracting key frames of surveillance video | |
| CN110769259A (en) | Image data compression method for tracking track content of video target | |
| CN109359530B (en) | Intelligent video monitoring method and device | |
| Kolekar et al. | Semantic event detection and classification in cricket video sequence | |
| Carbonneau et al. | Real-time visual play-break detection in sport events using a context descriptor | |
| Ouyang et al. | The comparison and analysis of extracting video key frame | |
| Premaratne et al. | Structural approach for event resolution in cricket videos | |
| CN115941997A (en) | Fragment-adaptive surveillance video concentration method | |
| Pan et al. | NTU TRECVID-2007 fast rushes summarization system | |
| Tonge et al. | A Novel Approach for Static Video Content Summarization using Shot Segmentation and k-means Clustering | |
| Broilo et al. | Unsupervised anchorpersons differentiation in news video | |
| Han et al. | A general framework for automatic on-line replay detection in sports video | |
| Tong et al. | A three-level scheme for real-time ball tracking | |
| Li et al. | Multiple object tracking using motion vectors from compressed video |
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 | ||
| CP03 | Change of name, title or address |
Address after: 100088 floor 15, block A, Haidian District urban construction, Beijing. Patentee after: China Huarong Technology Group Limited Address before: 100088 Haidian District, Beijing, North Taiping Road 18, city building A block 15. Patentee before: CHINA HUA RONG HOLDINGS CORPORATION LTD. |
|
| CP03 | Change of name, title or address | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20210310 Address after: Floor 15, block a, 18 Beitaipingzhuang Road, Haidian District, Beijing Patentee after: HUARONG TECHNOLOGY Co.,Ltd. Address before: 100088 floor 15, block A, Haidian District urban construction, Beijing. Patentee before: CHINA HUARONG TECHNOLOGY GROUP Co.,Ltd. |
|
| TR01 | Transfer of patent right |