CN103605652A - Video retrieval and browsing method and device based on object zone bits - Google Patents
Video retrieval and browsing method and device based on object zone bits Download PDFInfo
- Publication number
- CN103605652A CN103605652A CN201310388597.7A CN201310388597A CN103605652A CN 103605652 A CN103605652 A CN 103605652A CN 201310388597 A CN201310388597 A CN 201310388597A CN 103605652 A CN103605652 A CN 103605652A
- Authority
- CN
- China
- Prior art keywords
- video
- module
- retrieval
- concentrated
- flag position
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/74—Browsing; Visualisation therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/78—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/783—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
- G06F16/7847—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using low-level visual features of the video content
- G06F16/786—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using low-level visual features of the video content using motion, e.g. object motion or camera motion
Abstract
The invention relates to a video retrieval and browsing method based on object zone bits. The method includes the following steps that S1 a video concentration module conducts concentration to generate a concentrated video A based on an original video stored in a storage module and stores the video A into the storage module; S2 an object zone bit module acquires an object area zone bit and an object mapping zone bit based on the original video in the storage module in the process that the video concentration module generates the concentrated video A in the step S1; S3 a retrieval module searches for video segments meeting the requirement; S4 the video concentration module conducts concentration on the searched video segments to generate a concentrated video B; S5 a browse module conducts concentration and browsing on the concentrated video B generated in the step S4. A video retrieval and browsing device based on the object zone bits is further provided. By means of the method and device, high-efficiency retrieval and fast browsing of monitored videos can be achieved on the basis that the video code monitoring efficiency is not affected.
Description
Technical field
The invention relates to a kind of video frequency searching and method and the device browsed, particularly about a kind of video frequency searching based on object flag position and method and the device browsed.
Background technology
In the past few years, digital video monitoring has become a very important application.The technology wherein relating to, comprises the video surveillance applications requisite Video coding of institute and browser technology, has obtained increasing concern.Aspect video monitoring, most of videos are all caught by fixing camera; Even if for on-fixed camera, most of video segments also remain from fixed viewpoint and fixed position in the set time and record.Therefore, video technique is by utilizing this characteristic to provide a scheme more efficiently for video monitoring.
In monitor video field of storage, the adaptability that a lot of scholars are devoted to improve storage efficiency and increase code stream.For the monitor video scalable coding technology towards different channels, different terminals, different viewing demand, conduct in-depth research too.Yet traditional scalable coding technology has only solved the problem that flexible time domain, spatial domain are scalable and quality scalable is browsed based on frame.And how to retrieve more easily and to browse this problem of monitor video still unresolved to a certain extent.
Therefore, video frequency searching and video tour technology are progressively valued by the people, and have been applied in video monitoring system.Video summarization technique (A.Rav-Acha, Y.Pritch, and S.Peleg.Making a long video short:Dynamic video synopsis.In CVPR ' 06, volume1, pages435 – 441,2006.) summary as longer video makes people to access multitude of video content in the shorter time.Therefore, this is a very effective monitored video browse method.Unlike traditional video summarization technique, the dynamic video of video frequency abstract is concentrated, can be by show that the original object motion occurring when different in original video comes when retaining maximum useful content short-sighted frequency of original video boil down to simultaneously.Further, in order to meet, video is processed in real time and the demand of fast browsing, plum is blue or green waits people to propose concentrated online processing framework (the Feng S of video, Lei Z, Yi D, et al.Online content-aware video condensation[C] //Computer Vision and Pattern Recognition (CVPR), 2012IEEE Conference on.IEEE, 2012:2082-2087.).And telescopic video concentration technique based on details is suggested and give prominence to video content and improve storage efficiency (Wang S for laddering, Yang J, Zhao Y, et al.A surveillance video analysis and storage scheme for scalable synopsis browsing[C] //Computer Vision Workshops (ICCV Workshops), 2011IEEE International Conference on.IEEE, 2011:1947-1954.).
In addition, in the patent of invention that is 201110346398.0 at application number, introduced a kind of video coding-decoding method and device of supporting that video scalability is browsed, it relates to object flag position, for the mark to original video encoding and decoding video data, and, also related to video concentration technique, for the fast browsing of monitor video.Area information and the mapping relations of Moving Objects between original video and concentrated video of the described enough effective expression Moving Objects of object flag potential energy.In simple terms, object flag position comprises subject area zone bit and object map zone bit, their specifying information is: 1) subject area zone bit, first through video analysis, obtain the Pixel-level mask of Moving Objects, and generate a boundary rectangle frame based on mask, obtain boundary rectangle frame information, then the macroblock partitions information in object mask information and monitored video compression territory code stream is merged, utilize the piece division information in code stream Moving Objects to be carried out to the mark of sub-block level degree of accuracy; 2) object map zone bit, the Moving Objects that concentrated video can occur different time sections in original video is presented in same frame, therefore object map zone bit is designed to record the mapping relations of Moving Objects between original video and concentrated video, mark Moving Objects in original video which frame in concentrated video show.
With respect to video tour technology, the research of merging the video tour technology of video retrieval technology will be lacked.But, when video retrieval technology is applied to video monitoring system, can provide monitored video browse effect more targetedly to people, improve the service efficiency of monitor video.
Summary of the invention
The method and apparatus that the object of the present invention is to provide a kind of video frequency searching based on object flag position and browse, is not affecting on the basis of monitor video code efficiency, realizes efficient retrieval and the fast browsing of monitor video.
The present invention realizes by the following technical solutions.A kind of method that the invention provides video frequency searching based on object flag position and browse, comprises the following steps: S1, and the original video of the concentrated module of video based on storing in memory module concentrates and generates concentrated video A, and deposits memory module in; S2, the concentrated module of the video of object flag position module in step S1 generates in the process of concentrated video A, and the original video based in memory module, obtains subject area zone bit and object map zone bit, and deposits described zone bit information in memory module; S3, retrieval module retrieves satisfactory video segment based on object flag position; S4, the concentrated module of video concentrates and generates concentrated video B the video segment having retrieved, and deposits memory module in; And S5, browse module and the concentrated video B generating in the concentrated video A generating in step S1 or step S4 is carried out to video is concentrated to be browsed.
The present invention also can be applied to the following technical measures to achieve further.
Preferably, the aforesaid video frequency searching based on object flag position and the method for browsing, wherein in step S2, the subject area zone bit obtaining is the area information of Moving Objects in original video, and obtain object map zone bit, is the correspondence mappings information of Moving Objects between original video and concentrated video.
Preferably, the aforesaid video frequency searching based on object flag position and the method for browsing, wherein in step S2, when Moving Objects is mapped to concentrated video A, the priority of this Moving Objects when video frequency searching can be high, and in step S3, retrieval module is only retrieved the high Moving Objects of priority, thereby when guaranteeing retrieval effectiveness, further reduced data volume to be retrieved.
Preferably, the aforesaid video frequency searching based on object flag position and the method for browsing, wherein, in step S2, together with above-mentioned subject area zone bit, object map zone bit can also being organized in original video, make original video become the monitor video based on object flag position.
Preferably, the aforesaid video frequency searching based on object flag position and the method for browsing, wherein in step S3, retrieval module can be selected the mode of concrete retrieval, analyze the specifying information of the moving target in target to be retrieved and memory module, and contrast the matching degree of the information of the moving target in target to be retrieved and memory module, and then retrieve satisfactory video segment.
Preferably, the aforesaid video frequency searching based on object flag position and the method for browsing, wherein, in step S3, the mode that can retrieve comprises sample retrieval, color retrieval, shape retrieval and trajectory retrieval.
Preferably, the aforesaid video frequency searching based on object flag position and the method for browsing, wherein in step S5, described browse module and can also carry out scalable concentrated browsing.
The present invention also realizes by the following technical solutions.The device that the present invention also provides a kind of video frequency searching based on object flag position and browses, comprising: object flag position module, video concentrate module, memory module, retrieval module and browse module; This object flag position module is connecting the concentrated module of video and memory module, for carrying out mark to generating the Moving Objects region of original video of concentrated video and the scheme that is mapped to concentrated video thereof, to obtain subject area zone bit and object map zone bit information, and deposited in memory module; The concentrated module of this video is connecting memory module, concentrated for original video or the video segment that retrieved being carried out to video, and deposits the concentrated video generating in memory module; This retrieval module is connecting memory module, can carry out the retrieval of different modes, for retrieving satisfactory video segment; This is browsed module and is connecting memory module, can retrieve concentrated browsing.
The present invention also can be applied to the following technical measures to achieve further.
Preferably, the aforesaid video frequency searching based on object flag position and the device of browsing, wherein this object flag position module comprises subject area zone bit submodule and object map zone bit submodule, this subject area zone bit submodule is for obtaining the area information of original video Moving Objects, and this object map zone bit submodule is for obtaining the correspondence mappings information of Moving Objects between original video and concentrated video.
Preferably, the aforesaid video frequency searching based on object flag position and the device of browsing, wherein this object flag position module also comprises video compiling module, together with above-mentioned subject area zone bit, object map zone bit are organized in original video, make original video become the monitor video based on object flag position
Preferably, the aforesaid video frequency searching based on object flag position and the device of browsing, wherein this memory module is mainly used in storing original video, monitor video, retrieval threshold based on subject area zone bit and object map zone bit and the video segment having retrieved.
Preferably, the aforesaid video frequency searching based on object flag position and the device of browsing, the retrieval threshold of wherein storing in this memory module both can be set in advance, can input in real time again, and according to the difference of retrieval mode, each retrieval threshold is also different simultaneously.
Preferably, the aforesaid video frequency searching based on object flag position and the device of browsing, wherein this retrieval module comprises characteristics analysis module, comparative analysis module and load module; This load module is connecting characteristics analysis module, and this characteristics analysis module is connecting again comparative analysis module and memory module, and comparative analysis module is connecting memory module; This load module is for selecting the mode of concrete retrieval, this characteristics analysis module is according to the retrieval mode of having selected, analyze the specifying information of the moving target in target to be retrieved and memory module, similarity or the matching degree of the information of the moving target in this comparative analysis module comparative analysis target to be retrieved and memory module.
Preferably, the aforesaid video frequency searching based on object flag position and the device of browsing, wherein the mode of this retrieval comprises sample retrieval, color retrieval, shape retrieval and trajectory retrieval.
By technique scheme, the device and method that the present invention is based on the video frequency searching of object flag position and browse at least has following advantages and beneficial effect: by the present invention, can not affect on the basis of monitor video code efficiency, realize efficient retrieval and the fast browsing of monitor video.
Above-mentioned explanation is only the general introduction of technical solution of the present invention, in order to better understand technological means of the present invention, and can be implemented according to the content of instructions, and for above and other object of the present invention, feature and advantage can be become apparent, below especially exemplified by preferred embodiment, and coordinate accompanying drawing, be described in detail as follows.
Accompanying drawing explanation
Fig. 1: the flow chart of steps of the video frequency searching based on object flag position of the present invention and the method for browsing.
Fig. 2: the block scheme of the video frequency searching based on object flag position of the present invention and the device of browsing.
Fig. 3: the block scheme of retrieval module of the present invention.
Fig. 4 A, Fig. 4 B, Fig. 4 C and Fig. 4 D: the subjective effect schematic diagram after four kinds of retrieval modes of the present invention.
[main element symbol description]
1: object flag position module 2: video concentrates module
3: memory module 4: retrieval module
5: browse module 11: subject area zone bit submodule
12: object map zone bit submodule
13: video compiling module 41: characteristics analysis module
42: comparative analysis module 43: load module
A, B: concentrated video
Embodiment
For further setting forth the present invention, reach technological means and the effect that predetermined goal of the invention is taked, below in conjunction with accompanying drawing and preferred embodiment, embodiment, structure, feature and effect thereof to a kind of video frequency searching based on object flag position proposing according to the present invention with the device and method of browsing, be described in detail as follows.
Consulting shown in Fig. 2, is the block scheme of the video frequency searching based on object flag position of the present invention with the device of browsing.This device comprises object flag position module 1, video concentrated module 2, memory module 3, retrieval module 4 and browses module 5.
This object flag position module 1 is connecting the concentrated module 2 of video and memory module 3, for carrying out mark to generating the Moving Objects region of original video of concentrated video and the scheme that is mapped to concentrated video thereof, comprise subject area zone bit submodule 11 and object map zone bit submodule 12.This subject area zone bit submodule 11 is for obtaining the area information of original video Moving Objects, this area information has marked the region of Moving Objects in original video, it is subject area zone bit, and deposited in memory module 3, these can obtain texture, color, shape and the movement locus of Moving Objects in conjunction with original video.This object map zone bit submodule 12 is for obtaining the correspondence mappings information of Moving Objects between original video and concentrated video, i.e. object map zone bit, and deposited in memory module 3.In addition, this object flag position module 1 also comprises video compiling module 13, together with above-mentioned subject area zone bit is organized in original video with object map zone bit, make original video become the monitor video based on object flag position, and the monitor video based on object flag position deposits memory module 3 in by this, follow-up quick-searching and concentrated browsing have been beneficial to.
In the present embodiment, if the Moving Objects in original video is not mapped to concentrated video A, the significance level that can be understood as this Moving Objects is lower, its priority when video frequency searching is also lower like this, otherwise, when Moving Objects is mapped, the priority of this Moving Objects when video frequency searching is also higher.It should be noted that the present invention only retrieves the high subject area of priority, thereby when guaranteeing retrieval effectiveness, further reduced data volume to be retrieved.
The concentrated module 2 of this video is connecting memory module 3, both can be used for that original monitor video is carried out to video and has concentrated, and the video segment also can be used for having retrieved concentrates.Video is concentrated be one for the brand-new method of abstracting of monitor video fast browsing, thereby it effectively reduces the time-space domain redundancy while browsing original video by the object extracting from different frame is put together.Concrete video method for concentration can be referring to Publication about Document: A.Rav-Acha, Y.Pritch, and S.Peleg.Making a long video short:Dynamic video synopsis.In CVPR ' 06, volume1, pages435 – 441,2006.
This memory module 3 is mainly used in the video segment of storing original video, the monitor video based on object flag position (this object flag position refers to subject area zone bit and object map zone bit), retrieval threshold and having retrieved, wherein this retrieval threshold both can be set in advance, can be inputted in real time by the load module of retrieval module again.Specifically, according to the subject area zone bit in memory module 3, can obtain the Moving Objects area information in original video, and according to the object map zone bit in memory module 3, can obtain the priority of subject area and generate the mapping scheme that concentrates video A.
Consulting shown in Fig. 3, is the block scheme of retrieval module of the present invention.This retrieval module 4 is connecting memory module 3, comprises characteristics analysis module 41, comparative analysis module 42 and load module 43; This load module 43 is connecting characteristics analysis module 41, and this characteristics analysis module 41 is connecting again comparative analysis module 42, and comparative analysis module 42 is connecting memory module 3.This load module 43 is for selecting the mode of concrete retrieval.This characteristics analysis module 41, according to the retrieval mode of having selected, is analyzed the specifying information of the moving target in target to be retrieved and memory module 3.Whether this comparative analysis module 42, for contrasting the similarity of information of moving target of target to be retrieved and memory module 3 or matching degree etc., reaches threshold value.When meeting the requirements, can deposit the video segment having retrieved in memory module 3.
In the present embodiment, the retrieval mode relating generally to is sample retrieval, color retrieval, shape retrieval and trajectory retrieval.To be described in detail above-mentioned four kinds of retrieval modes below:
Sample retrieval: by load module 43, user directly inputs the sample picture of a target to be retrieved, characteristics analysis module 41 receives this sample picture subsequently, and extract textural characteristics (the concrete extracting method of sample picture, referring to document: P.Wu, B.S.Manjunath, S.D.Newsam, and H.D.Shin, " A texture descriptor for image retrieval and browsing, " in Computer Vision and Pattern Recognition Workshop, Fort Collins, CO, Jun.1999, pp.3 – 7.), also can extract the textural characteristics of each moving target in memory module 3 simultaneously, and be sent to comparative analysis module 42, the textural characteristics of these comparative analysis module 42 comparative analysis sample pictures and the textural characteristics of each moving target in memory module 3, when the matching degree with some moving targets surpasses predefined threshold value (being more than or equal to 80%), represent to meet the requirements, and deposit this moving target in memory module 3.
Consult shown in Fig. 4 A, by load module 43, user selects directly to input the picture of a car after sample retrieval, characteristics analysis module 41 receives this car picture, and extracting the textural characteristics of car picture, the textural characteristics of the 42 comparative analysis car pictures of comparative analysis module subsequently and the textural characteristics of each moving target in memory module 3, when the matching degree with some moving targets surpasses 80%, represent to meet the requirements, and deposit this Moving Objects in memory module 3.Fig. 4 A is the subjective effect schematic diagram that utilizes the car that sample retrieves.
Color retrieval: color is Moving Objects one of feature the most intuitively in monitor video, in the present invention, can pass through simple color histogram (specifically referring to document: B.S.Manjunath, J.-R.Ohm, V.V.Vasudevan, and A.Yamada, " Color and Texture Descriptors; " IEEE Trans.Circuits and Systems for Video Technology, 11 (6): 703 – 715,2001.) coupling is carried out the color retrieval of Moving Objects.In the present invention, the monitor video decoding has been transformed into RGB color space.By load module 43, color retrieval is carried out in selection, characteristics analysis module 41 is analyzed the color of target to be retrieved subsequently, and the matching degree of the RGB color space of each moving target in comparative analysis module 42 comparative analysis target to be retrieved and memory module 3, when the matching degree mean value between each Color Channel is greater than certain value (threshold value is 70%), be defined as mutual coupling, otherwise represent not mate.Finally, the Moving Objects of selecting coupling deposits memory module 3 in.
Consult shown in Fig. 4 B, by load module 43, user selects in the palette of system, to input selected color after color retrieval, characteristics analysis module 41 receives after selected color, the matching degree of three color spaces of each moving target in comparative analysis module 42 comparative analysis input colors and memory module 3, matching degree mean value between each Color Channel is greater than 70%, is defined as mutually mating and depositing in memory module 3.Fig. 4 A is the subjective effect schematic diagram that utilizes the car that color retrieves.
Shape retrieval: shape retrieval is mainly that the profile diagram of retrieving or directly input destination object by the profile diagram of Freehandhand-drawing destination object is retrieved.By load module 43, select to carry out shape retrieval, characteristics analysis module 41 is analyzed the shape of target to be retrieved subsequently, and the matching degree of the shape of each moving target in comparative analysis module 42 comparative analysis target to be retrieved and memory module 3.Finally, the Moving Objects of selecting coupling deposits memory module 3 in.
Specifically, in the present invention, the shape of the moving target of each marked area can represent with 7 degree of freedom eigenvector.(method that specifically generates is referring to M.Hu, " Visual pattern recognition by moment invariants, " IRE Trans.Inform.Theory, and vol.IT-8, no.2, pp.179 – 182, Feb.1962.)
If the shape representation of moving target A is:
and the shape representation of moving target B is:
similarity between moving target A and the shape of B can be calculated by following formula:
When the similarity between moving target A and the shape of B surpasses threshold value 60%, just think that moving target A and B match.
Consult shown in Fig. 4 C, by load module 43, user selects to carry out shape and retrieves the profile diagram that also Freehandhand-drawing goes out a car, characteristics analysis module 41 is analyzed this profile diagram information extraction, the matching degree of the shape of each moving target in 42 these profile diagrams of comparative analysis of comparative analysis module subsequently and memory module 3, when similarity surpasses 60%, think that the match is successful, deposit memory module 3 in.Fig. 4 C is the subjective effect schematic diagram that utilizes the car that shape retrieves.
In fact, the retrieval based on Moving Objects shape can search out the Moving Objects with Representative character easily, and as pedestrian, vehicle two type games objects have obvious shape difference.But, the shape of Moving Objects is meeting and the direction of motion of Moving Objects and the Angular correlation of monitoring camera directly, may there is change of shape clearly in same non-rigid motion object, these factors also make shape retrieval have some limitations in video.
Trajectory retrieval: except above-mentioned static nature can be retrieved, the present invention can also carry out the trajectory retrieval of Moving Objects, the feature matching method of trajectory retrieval and other modes is different, because the track of Freehandhand-drawing has time domain specification, so the matching process of track does not need Moving Objects to mate frame by frame, but the set of integral body point is retrieved to coupling.(can be referring to document: JunWei Hsieh, Shang-Li Yu and Yung-Sheng Chen, Motion-based video retrieval by trajectory matching, IEEE Transactions on Circuits and Systems for Video Technology, Vol.16, No.3,396-409,2006.)
Concrete performing step is as follows:
If the trajectory table of Moving Objects is shown:
M wherein
kthe track that represents k Moving Objects is counted, and N is Moving Objects sum to be retrieved.The trajectory table of user's Freehandhand-drawing is shown:
M wherein
rthe coordinate that represents retrieval track is counted.
General, the coordinate of the retrieval track of the input m that counts
rwith the track of the Moving Objects m that counts
kbetween have significant difference.Therefore,, if calculate two distances between track each point, so just very possible erroneous judgement, misses the Moving Objects that originally can mate.Thus, we have carried out normalization to the track of target trajectory and target to be retrieved:
If m
k> m
r, calculate scale-up factor Ra=m
k/ m
r, by T
kcoefficients R a abandons and obtains T ' in proportion
k, the similarity of then carrying out is between points calculated.Same, if m
k< m
r, also can adopt similar method, carry out the track similarity of carrying out again between track after normalization of counting and calculate.
When the similarity between track surpasses threshold value 50%, just think and mate with searched targets, and deposit memory module in.
Consult shown in Fig. 4 D, by load module 43, user selects to carry out trajectory retrieval and Freehandhand-drawing goes out Moving Objects track to be retrieved, characteristics analysis module 41 is analyzed this Moving Objects track, the trace information of each moving target in 42 these tracks of comparative analysis of comparative analysis module subsequently and memory module 3, selection matching degree is greater than 50% moving target as successful matching result, and deposits memory module 3 in.Fig. 4 D is the subjective effect schematic diagram that utilizes the car that trajectory retrieval retrieves.
This is browsed module 5 and is connecting memory module 4, can carry out concentrated the browsing of retrieval of result for retrieval.Specifically, the browse mode relating in the present invention is browsed for retrieval is concentrated, and it is mainly on the basis of retrieval, for the video segment that meets retrieval requirement, by the concentrated module 2 of video, generate concentrated video, and browse, thereby make the concentrated video of new retrieval more pointed.In fact, this is browsed module 5 and also can carry out scalable concentrated browsing.In order to realize scalable concentrated browsing, this is browsed module 5 and has realized a kind of scalable method for reconstructing based on object map zone bit, the method can be adjusted the mapping relations of Moving Objects, and reconstructs the multiple concentrated video of different length, different densities and different fidelitys.And user can, according to watching demand, arrange scalable progression and reconstruct oneself satisfied concentrated video.Specifically can be referring to document: Wang, S., Yang, J., and Stan Z.Li, A Surveillance Video Analysis and Storage Scheme for Scalable Synopsis Browsing.ICCV Workshop on Visual Surveillance, November, 2011.
To the video frequency searching based on object flag position of the present invention and the method for browsing be described in detail below, it mainly comprises the following steps (consulting shown in Fig. 1):
S1, the original video of the concentrated module 2 of video based on storage in memory module 3 carries out video analysis, obtain the area information of Moving Objects, and then it is concentrated to carry out video, obtain the mapping scheme of Moving Objects in original video and concentrated video, generate concentrated video A, and will concentrate video A and deposit memory module 3 in;
S2, the concentrated module of the video of object flag position module 1 in step S1 generates in the process of concentrated video A, area information and the mapping scheme of the Moving Objects based on obtaining, formation object area flag position and object map zone bit, and deposit described zone bit information in memory module 3;
Further use the video compiling module 13 in object flag position module 1, together with subject area zone bit in this step S2, object map zone bit are organized in original video, make original video become through object flag position mark the monitor video of Moving Objects region and Moving Objects priority, and the monitor video based on object flag position deposits memory module 3 in by this, be beneficial to follow-up quick-searching and retrieve concentrated browsing;
S3, load module 43 by retrieval module 4, select after the mode of concrete retrieval, the specifying information of the moving target in characteristics analysis module 41 analysis targets to be retrieved and memory module 3, and by comparative analysis module 42, contrast the similarity of information of the moving target in target to be retrieved and memory module 3 or matching degree etc., and deposit the video segment having retrieved in memory module 3; Wherein, can carry out the retrieval of different modes here, for example sample retrieval, color retrieval, shape retrieval and trajectory retrieval;
S4, the concentrated 2 pairs of video segments that retrieved of module of video concentrate and become concentrated video B, and deposit memory module 3 in;
S5, browses the concentrated video B generating in the concentrated video A that generates in 5 couples of step S1 of module or step S4 and show, and browse mode is scalable concentrated concentrated the browsing of browsing or retrieve.
By the video frequency searching based on object flag position of the present invention and the device and method of browsing, can not affect on the basis of monitor video code efficiency, realize efficient retrieval and the fast browsing of monitor video.
The above, it is only preferred embodiment of the present invention, not the present invention is done to any pro forma restriction, although the present invention discloses as above with preferred embodiment, yet not in order to limit the present invention, any those skilled in the art, do not departing within the scope of technical solution of the present invention, when can utilizing the technology contents of above-mentioned announcement to make a little change or being modified to the equivalent embodiment of equivalent variations, in every case be the content that does not depart from technical solution of the present invention, any simple modification of above embodiment being done according to technical spirit of the present invention, equivalent variations and modification, all still belong in the scope of technical solution of the present invention.
Claims (14)
1. the video frequency searching based on object flag position and the method browsed, is characterized in that it comprises the following steps:
S1, the original video of the concentrated module of video based on storing in memory module concentrates and generates concentrated video A, and deposits memory module in;
S2, the concentrated module of the video of object flag position module in step S1 generates in the process of concentrated video A, and the original video based in memory module, obtains subject area zone bit and object map zone bit, and deposits described zone bit information in memory module;
S3, retrieval module retrieves satisfactory video segment based on object flag position, and deposits memory module in;
S4, the concentrated module of video concentrates and generates concentrated video B the video segment having retrieved, and deposits memory module in; And
S5, browses module and the concentrated video B generating in the concentrated video A generating in step S1 or step S4 is carried out to video is concentrated to be browsed.
2. the video frequency searching based on object flag position according to claim 1 and the method browsed, it is characterized in that wherein in step S2, the subject area zone bit obtaining is the area information of Moving Objects in original video, and the object map zone bit obtaining is the correspondence mappings information of Moving Objects between original video and concentrated video A.
3. the video frequency searching based on object flag position according to claim 2 and the method browsed, it is characterized in that wherein in step S2, when Moving Objects is mapped to concentrated video A, the priority of this Moving Objects when video frequency searching can be high, and in step S3, retrieval module is only retrieved the high Moving Objects of priority, thereby when guaranteeing retrieval effectiveness, has further reduced data volume to be retrieved.
4. the video frequency searching based on object flag position according to claim 1 and the method browsed, it is characterized in that wherein in step S2, together with above-mentioned subject area zone bit, object map zone bit can also being organized in original video, make original video become the monitor video based on object flag position.
5. the video frequency searching based on object flag position according to claim 1 and the method browsed, it is characterized in that wherein in step S3, retrieval module can be selected the mode of concrete retrieval, analyze the specifying information of the moving target in target to be retrieved and memory module, and contrast the matching degree of the information of the moving target in target to be retrieved and memory module, and then retrieve satisfactory video segment.
6. the video frequency searching based on object flag position according to claim 5 and the method browsed, is characterized in that wherein, in step S3, the mode that can retrieve comprises sample retrieval, color retrieval, shape retrieval and trajectory retrieval.
7. the video frequency searching based on object flag position according to claim 1 and the method browsed, is characterized in that wherein in step S5, described browse module and can also carry out scalable concentrated browsing.
8. the video frequency searching based on object flag position and the device browsed, is characterized in that comprising:
Object flag position module, video concentrate module, memory module, retrieval module and browse module;
This object flag position module is connecting the concentrated module of video and memory module, for carrying out mark to generating the Moving Objects region of original video of concentrated video and the scheme that is mapped to concentrated video thereof, to obtain subject area zone bit and object map zone bit information, and deposited in memory module;
The concentrated module of this video is connecting memory module, concentrated for original video or the video segment that retrieved being carried out to video, and deposits the concentrated video generating in memory module; This retrieval module is connecting memory module, can carry out the retrieval of different modes, for retrieving satisfactory video segment, and is deposited in memory module;
This is browsed module and is connecting memory module, can retrieve concentrated browsing.
9. the video frequency searching based on object flag position according to claim 8 and the device browsed, it is characterized in that wherein this object flag position module comprises subject area zone bit submodule and object map zone bit submodule, this subject area zone bit submodule is for obtaining the area information of original video Moving Objects, and this object map zone bit submodule is for obtaining the correspondence mappings information of Moving Objects between original video and concentrated video.
10. the video frequency searching based on object flag position according to claim 9 and the device browsed, it is characterized in that wherein this object flag position module also comprises video compiling module, together with above-mentioned subject area zone bit, object map zone bit are organized in original video, make original video become the monitor video based on object flag position.
11. video frequency searchings based on object flag position according to claim 8 and the device of browsing, is characterized in that wherein this memory module is mainly used in the video segment of storing original video, the monitor video based on object flag position, retrieval threshold and having retrieved.
12. video frequency searchings based on object flag position according to claim 11 and the device of browsing, it is characterized in that the retrieval threshold of storing in this memory module wherein both can set in advance, can input in real time again, according to the difference of retrieval mode, each retrieval threshold is also different simultaneously.
13. video frequency searchings based on object flag position according to claim 8 and the device of browsing, is characterized in that wherein this retrieval module comprises characteristics analysis module, comparative analysis module and load module; This load module is connecting characteristics analysis module, and this characteristics analysis module is connecting again comparative analysis module and memory module, and comparative analysis module is connecting memory module;
This load module is for selecting the mode of concrete retrieval, this characteristics analysis module is according to the retrieval mode of having selected, analyze the specifying information of the moving target in target to be retrieved and memory module, similarity or the matching degree of the information of the moving target in this comparative analysis module comparative analysis target to be retrieved and memory module.
14. video frequency searchings based on object flag position according to claim 8 and the device of browsing, is characterized in that the mode of wherein this retrieval comprises sample retrieval, color retrieval, shape retrieval and trajectory retrieval.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310388597.7A CN103605652B (en) | 2013-08-30 | 2013-08-30 | Video frequency searching and the method and apparatus browsed based on object flag position |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310388597.7A CN103605652B (en) | 2013-08-30 | 2013-08-30 | Video frequency searching and the method and apparatus browsed based on object flag position |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103605652A true CN103605652A (en) | 2014-02-26 |
| CN103605652B CN103605652B (en) | 2017-11-07 |
Family
ID=50123877
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310388597.7A Expired - Fee Related CN103605652B (en) | 2013-08-30 | 2013-08-30 | Video frequency searching and the method and apparatus browsed based on object flag position |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103605652B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103873864A (en) * | 2014-03-31 | 2014-06-18 | 江南大学 | Object flag bit efficient encoding method applied to video object retrieval |
| CN104077384A (en) * | 2014-06-27 | 2014-10-01 | 深圳市中兴移动通信有限公司 | Mobile terminal as well as method and device for realizing video indexing by using screen splitting function |
| CN106385562A (en) * | 2016-09-23 | 2017-02-08 | 浙江宇视科技有限公司 | Video abstract generation method, device and video monitoring system |
| CN106445985A (en) * | 2016-04-29 | 2017-02-22 | 上海交通大学 | Hand painted motion contour-based video search method and system |
| CN110557589A (en) * | 2018-05-30 | 2019-12-10 | 百度（美国）有限责任公司 | System and method for integrating recorded content |
| CN111159476A (en) * | 2019-12-11 | 2020-05-15 | 智慧眼科技股份有限公司 | Target object searching method and device, computer equipment and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102156707A (en) * | 2011-02-01 | 2011-08-17 | 刘中华 | Video abstract forming and searching method and system |
| CN102222104A (en) * | 2011-06-23 | 2011-10-19 | 华南理工大学 | Method for intelligently extracting video abstract based on time-space fusion |
| CN102395029A (en) * | 2011-11-05 | 2012-03-28 | 江苏物联网研究发展中心 | Video encoding and decoding method and device supporting retractable video browse |
| US8401252B2 (en) * | 2006-09-29 | 2013-03-19 | Google Inc. | Video retrieval system for human face content |
| CN103020275A (en) * | 2012-12-28 | 2013-04-03 | 苏州千视通信科技有限公司 | Video analysis method based on video abstraction and video retrieval |
| CN103150375A (en) * | 2013-03-11 | 2013-06-12 | 浙江捷尚视觉科技有限公司 | Quick video retrieval system and quick video retrieval method for video detection |
| CN103246922A (en) * | 2013-05-07 | 2013-08-14 | 重庆大学 | Method for video abstract generation |
-
2013
- 2013-08-30 CN CN201310388597.7A patent/CN103605652B/en not_active Expired - Fee Related
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8401252B2 (en) * | 2006-09-29 | 2013-03-19 | Google Inc. | Video retrieval system for human face content |
| CN102156707A (en) * | 2011-02-01 | 2011-08-17 | 刘中华 | Video abstract forming and searching method and system |
| CN102222104A (en) * | 2011-06-23 | 2011-10-19 | 华南理工大学 | Method for intelligently extracting video abstract based on time-space fusion |
| CN102395029A (en) * | 2011-11-05 | 2012-03-28 | 江苏物联网研究发展中心 | Video encoding and decoding method and device supporting retractable video browse |
| CN103020275A (en) * | 2012-12-28 | 2013-04-03 | 苏州千视通信科技有限公司 | Video analysis method based on video abstraction and video retrieval |
| CN103150375A (en) * | 2013-03-11 | 2013-06-12 | 浙江捷尚视觉科技有限公司 | Quick video retrieval system and quick video retrieval method for video detection |
| CN103246922A (en) * | 2013-05-07 | 2013-08-14 | 重庆大学 | Method for video abstract generation |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103873864A (en) * | 2014-03-31 | 2014-06-18 | 江南大学 | Object flag bit efficient encoding method applied to video object retrieval |
| CN104077384A (en) * | 2014-06-27 | 2014-10-01 | 深圳市中兴移动通信有限公司 | Mobile terminal as well as method and device for realizing video indexing by using screen splitting function |
| CN106445985A (en) * | 2016-04-29 | 2017-02-22 | 上海交通大学 | Hand painted motion contour-based video search method and system |
| CN106445985B (en) * | 2016-04-29 | 2019-09-03 | 上海交通大学 | Video retrieval method and system based on Freehandhand-drawing motion outline |
| CN106385562A (en) * | 2016-09-23 | 2017-02-08 | 浙江宇视科技有限公司 | Video abstract generation method, device and video monitoring system |
| CN106385562B (en) * | 2016-09-23 | 2019-06-11 | 浙江宇视科技有限公司 | A kind of video abstraction generating method, device and video monitoring system |
| CN110557589A (en) * | 2018-05-30 | 2019-12-10 | 百度（美国）有限责任公司 | System and method for integrating recorded content |
| CN111159476A (en) * | 2019-12-11 | 2020-05-15 | 智慧眼科技股份有限公司 | Target object searching method and device, computer equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103605652B (en) | 2017-11-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kuanar et al. | Video key frame extraction through dynamic Delaunay clustering with a structural constraint | |
| CN103605652A (en) | Video retrieval and browsing method and device based on object zone bits | |
| US8515933B2 (en) | Video search method, video search system, and method thereof for establishing video database | |
| Basavarajaiah et al. | Survey of compressed domain video summarization techniques | |
| Mendi et al. | Sports video summarization based on motion analysis | |
| US20090290752A1 (en) | Method for producing video signatures and identifying video clips | |
| Meng et al. | Object instance search in videos via spatio-temporal trajectory discovery | |
| Amiri et al. | Video shot boundary detection using generalized eigenvalue decomposition and Gaussian transition detection | |
| Ding et al. | Survsurf: human retrieval on large surveillance video data | |
| Song et al. | Joint key-frame extraction and object segmentation for content-based video analysis | |
| Thomas et al. | Perceptual synoptic view-based video retrieval using metadata | |
| Liu et al. | HVS-based perception-driven no-reference omnidirectional image quality assessment | |
| Liu et al. | Real-time video copy-location detection in large-scale repositories | |
| Garcia | Temporal aggregation of visual features for large-scale image-to-video retrieval | |
| HE et al. | Chronological video synopsis via events rearrangement optimization | |
| He et al. | Self-supervised video retrieval transformer network | |
| JP2002513487A (en) | Algorithms and systems for video search based on object-oriented content | |
| Joshi et al. | On fuzzy clustering and content based access to networked video databases | |
| US8666164B2 (en) | System and method for modeling a region segmented image | |
| Idris et al. | Image and video indexing using vector quantization | |
| Mallick et al. | Video retrieval using salient foreground region of motion vector based extracted keyframes and spatial pyramid matching | |
| Yang et al. | Content-based video retrieval (cbvr) system for cctv surveillance videos | |
| del Fabro et al. | Video scene detection based on recurring motion patterns | |
| CN108304453A (en) | A kind of determination method and device of video relevant search word | |
| Ye et al. | A surveillance video index and browsing system based on object flags and video synopsis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20171107 Termination date: 20200830 |