CN107578011A - The decision method and device of key frame of video - Google Patents
The decision method and device of key frame of video Download PDFInfo
- Publication number
- CN107578011A CN107578011A CN201710788529.8A CN201710788529A CN107578011A CN 107578011 A CN107578011 A CN 107578011A CN 201710788529 A CN201710788529 A CN 201710788529A CN 107578011 A CN107578011 A CN 107578011A
- Authority
- CN
- China
- Prior art keywords
- frame
- video
- local feature
- feature region
- key
- 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.)
- Pending
Links
Abstract
The invention provides a kind of decision method of key frame of video and device, is related to the technical field that video data is handled, and this method includes：The frame of video of target video is read frame by frame, by the first frame of video of target video, labeled as reference frame；The local feature region of reference frame is extracted, and local feature region is labeled as First partial characteristic point；Each subsequent frame of reference frame is obtained successively, extracts the local feature region of each subsequent frame；Judge whether the local feature region of each subsequent frame matches with First partial characteristic point；If not, subsequent frame is determined as to the key frame of target video.The decision method and device of key frame of video of the present invention, it is reference frame by choosing first frame of video, and extract the local feature region of reference frame, and the mode for being matched the local feature region of subsequent frame with the local feature region of reference frame, redundancy key frames can significantly be eliminated, meanwhile calculating process is also simplify, improve the efficiency of video analysis.
Description
Technical field
The present invention relates to the technical field of video data processing, a kind of decision method more particularly, to key frame of video and
Device.
Background technology
With the constantly improve of network infrastructure, the application of new network technology and the popularization of video capture equipment, depending on
Frequency total resources and people sharply increase to the demand of video resource.Intelligent dimension, automatic classification are carried out to video information,
Realize that the demands such as video content automatic detection, filter Video content retrieval are continuously increased.Key frame of video extraction is used as video
The basic technology of information processing, its performance directly affect the result of advanced video processing.
Existing key frame decision technology is in industry with having had relatively broad application, and different lives in life
Scene can select different key frame decision methods to be judged, can be selected such as in the case of video background is simple and be based on taking out
The key frame decision technology of sample and the key frame decision technology based on shot boundary, it can use and be based in the complicated video of background
Motion analysis and the key frame decision technology of cluster.
But because the determination of key frame will rely on changing for feature (such as color, texture, shape, the light stream) of video frame content
Become to determine, it is possible to situations below be present：Same object (a such as car) occurs in one scenario, but due to light
According to, angle and the difference of yardstick, it can reflect in frame of video feature and provide big difference (i.e. close or similar object exists
Very big difference can occur for the numerical value stored in machine), but they are same objects, therefore we should be determined as redundancy key
Frame, but conventional method does not have the arbitration functions of redundancy key frames typically, therefore redundancy key frames can be divided into key frame, lead
Cause key-frame extraction amount big, not only calculating process is complicated, also reduces the efficiency of video analysis.
The content of the invention
In view of this, it is an object of the invention to provide a kind of decision method of key frame of video and device, effectively to carry
Take the key frame of video.
In a first aspect, the embodiments of the invention provide a kind of decision method of key frame of video, including：Target is read frame by frame
The frame of video of video, by the first frame of video of target video, labeled as reference frame；The local feature region of reference frame is extracted, and will
Local feature region is labeled as First partial characteristic point；Each subsequent frame of reference frame is obtained successively, extracts the office of each subsequent frame
Portion's characteristic point；Judge whether the local feature region of each subsequent frame matches with First partial characteristic point；If not, subsequent frame is sentenced
It is set to the key frame of target video.
In preferred embodiments of the present invention, the above method also includes：After the first key frame of target video is determined,
Using key frame as reference frame, continue to judge the key frame of target video, until when reference frame is without subsequent frame terminating.
In preferred embodiments of the present invention, the above method includes：Local feature region is extracted using SURF algorithm, and sentenced
Whether the local feature region of disconnected each subsequent frame matches with above-mentioned First partial characteristic point.
In preferred embodiments of the present invention, local feature region and the First partial characteristic point of the above-mentioned each subsequent frame of judgement
Whether matching includes：Judge that the local feature region of each subsequent frame is set in advance with whether the similarity of First partial characteristic point is more than
Fixed threshold value；If it is, determining that subsequent frame matches with reference frame, and subsequent frame is labeled as redundancy key frames；If not, determine
Subsequent frame mismatches with reference frame.
In preferred embodiments of the present invention, the above method also includes：Target video is divided into multiple sub-goal videos；
The frame of video of each sub-goal video is read frame by frame, is judged with the key frame to each sub-goal video.
In preferred embodiments of the present invention, the above method also includes：Obtain the key frame of target video；To comprising
The target video for having key frame is retrieved, to retrieve the video image similar to target video.
Second aspect, the embodiments of the invention provide a kind of decision maker of key frame of video, including：With reference to frame flag mould
Block, it is reference frame by the first video frame indicia of target video for reading the frame of video of target video frame by frame；First partial is special
Sign point extraction module, First partial characteristic point is labeled as extracting the local feature region of reference frame, and by local feature region；The
Two local feature region extraction modules, for obtaining each subsequent frame of reference frame successively, extract the local feature of each subsequent frame
Point；Matching judgment module, for judging whether the local feature region of each subsequent frame matches with First partial characteristic point；Judge mould
Block, for when the judged result of matching judgment module is no, subsequent frame to be determined as to the key frame of target video.
In preferred embodiments of the present invention, said apparatus also includes：Reference frame chooses module, and target is determined for working as
After the first key frame of video, using key frame as reference frame, continue to judge the key frame of target video, until reference frame is without follow-up
Terminate during frame.
In preferred embodiments of the present invention, said apparatus also includes：Module is retrieved, for obtaining the key of target video
Frame；The target video for including key frame is retrieved, to retrieve the video image similar to target video.
The third aspect, the embodiments of the invention provide a kind of server, the server includes processor, memory, bus
And communication interface, processor, communication interface and memory are connected by bus；Memory is used for storage program；Processor, it is used for
The program of storage in memory is called by bus, performs above-mentioned first aspect methods described.
Fourth aspect, the embodiments of the invention provide a kind of computer-readable storage medium, for saving as above-mentioned Video Key
Computer software instructions used in the decision maker of frame, it includes the decision maker for being used for performing that above-mentioned aspect is key frame of video
Designed program.
The embodiment of the present invention brings following beneficial effect：
The decision method and device of a kind of key frame of video described in the embodiment of the present invention, by reading target video frame by frame
Frame of video, it is reference frame to choose first frame of video, can extract the local feature region of reference frame, and obtains reference frame successively
Each subsequent frame, the mode that the local feature region of subsequent frame is matched with the local feature region of reference frame, can be significantly
Ground eliminates redundancy key frames, meanwhile, calculating process is also simplify, improves the efficiency of video analysis.
Other features and advantages of the present invention will illustrate in the following description, also, partly become from specification
Obtain it is clear that or being understood by implementing the present invention.The purpose of the present invention and other advantages are in specification, claims
And specifically noted structure is realized and obtained in accompanying drawing.
To enable the above objects, features and advantages of the present invention to become apparent, preferred embodiment cited below particularly, and coordinate
Appended accompanying drawing, is described in detail below.
Brief description of the drawings
, below will be to specific in order to illustrate more clearly of the specific embodiment of the invention or technical scheme of the prior art
The required accompanying drawing used is briefly described in embodiment or description of the prior art, it should be apparent that, in describing below
Accompanying drawing is some embodiments of the present invention, for those skilled in the art, on the premise of not paying creative work,
Other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 is a kind of flow chart of the decision method of key frame of video provided in an embodiment of the present invention；
Fig. 2 is the flow chart of the decision method of another key frame of video provided in an embodiment of the present invention；
Fig. 3 is a kind of structural representation of the decision maker of key frame of video provided in an embodiment of the present invention；
Fig. 4 is the structural representation of the decision maker of another key frame of video provided in an embodiment of the present invention；
Fig. 5 is a kind of structural representation of server provided in an embodiment of the present invention.
Embodiment
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with accompanying drawing to the present invention
Technical scheme be clearly and completely described, it is clear that described embodiment is part of the embodiment of the present invention, rather than
Whole embodiments.Based on the embodiment in the present invention, those skilled in the art institute under the premise of creative work is not made
The every other embodiment obtained, belongs to the scope of protection of the invention.
Key frame is the representative frame of video lens, refers to select a frame or some representatives in the relatively independent camera lens of meaning
The frame of video of property represents the entire content of camera lens.Because video is made up of thousands of frame, t in video data stream
Picture frame and the t+1 even picture frame at t+2 moment on visual signature and content if difference less (we are referred to as redundancy key
Frame) when, we are not necessarily to pay close attention to redundancy key frames when browsing video, can so greatly save the time of user, effectively
The efficiency of video analysis is improved, is widely used in fields such as security protection, public security, digital librarys, based on this, the present invention is implemented
Example provides the decision method and device of a kind of key frame of video, to reduce influence of the redundancy key frames in video analysis.
For ease of understanding the present embodiment, sentencing to a kind of key frame of video disclosed in the embodiment of the present invention first
The method of determining describes in detail.
Embodiment one：
The embodiments of the invention provide a kind of decision method of key frame of video, a kind of key frame of video as shown in Figure 1
Decision method flow chart, comprise the following steps：
Step S102, the frame of video of target video is read frame by frame, is reference frame by the first video frame indicia of target video；
Step S104, the local feature region of reference frame is extracted, and local feature region is labeled as First partial characteristic point；
Step S106, each subsequent frame of reference frame is obtained successively, extract the local feature region of each subsequent frame；
Step S108, judges whether the local feature region of each subsequent frame matches with First partial characteristic point；
Step S110, if not, subsequent frame to be determined as to the key frame of target video.
The decision method of a kind of key frame of video described in the embodiment of the present invention, by the video for reading target video frame by frame
Frame, it is reference frame to choose first frame of video, can extract the local feature region of reference frame, and obtain successively reference frame it is each after
Continuous frame, the mode that the local feature region of subsequent frame is matched with the local feature region of reference frame, can significantly be eliminated
Redundancy key frames, meanwhile, calculating process is also simplify, improves the efficiency of video analysis.
Embodiment two：
In view of that can include substantial amounts of frame of video in usual target video, therefore, the above method is also included when by above-mentioned
After method described in embodiment determines first key frame, using the first key frame as reference frame, continue to judge target video
Key frame, until terminating when reference frame is without subsequent frame.
Based on this, on the basis of above-described embodiment, the embodiment of the present invention additionally provides sentencing for another key frame of video
Determine method, the flow chart of the decision method of another key frame of video as shown in Figure 2, comprise the following steps：
Step S202, the frame of video of target video is read frame by frame, is reference frame by the first video frame indicia of target video；
Step S204, the local feature region of reference frame is extracted, and local feature region is labeled as First partial characteristic point；
Step S206, the subsequent frame of the reference frame is obtained, extract the local feature region of the subsequent frame；
Step S208, judges whether the local feature region of the subsequent frame matches with First partial characteristic point；If it is, perform
Step S210；If not, perform step S216；
Step S210, the subsequent frame is determined as to the key frame of target video；
Step S212, judges whether the key frame has subsequent frame；If it is, perform step S214；If not, terminate；
Step S214, it is reference frame by the key frame marker；Return to step S204；
Step S216, the subsequent frame is determined as to the redundancy key frames of target video；
Step S218, judges whether the redundancy key frames have subsequent frame；If it is, return to step S206；If not, knot
Beam.
Preferably, local feature region is the local expression of frame of video feature, embodies the local feature having on two field picture,
It is especially suitable for carrying out the operation such as images match, and to the interference effect very little of environment, such as the changing, be rotationally-varying of illumination,
Angle change, influence of noise etc..Therefore, the mode of matching local feature region is taken in the embodiment of the present invention is to judge frame of video
No is key frame.
Generally, conventional local feature region detection algorithm has SIFT algorithms (Scale-invariant feature
Transform, SIFT, Scale invariant features transform), (Speeded Up Robust Features, SURF, add SURF algorithm
Fast robust feature), FAST feature point detection algorithms etc..It is real in Practical Project because SIFT and FAST algorithm amounts of calculation are larger
When property is restricted, therefore, after the embodiment of the present invention is preferably to extract local feature region using SURF algorithm, and judgement is each
Whether the local feature region of continuous frame matches with First partial characteristic point.
During specific implementation, step that whether the above-mentioned local feature region for judging each subsequent frame matches with First partial characteristic point
Suddenly following process can be included：(1) judge the local feature region of each subsequent frame and First partial characteristic point similarity whether
More than threshold value set in advance；(2) if it is, determining that subsequent frame matches with reference frame, and subsequent frame is crucial labeled as redundancy
Frame；(3) if not, determining that subsequent frame mismatches with reference frame, now the subsequent frame is key frame.
For the ease of understanding the embodiment of the present invention, the embodiment of the present invention is also local special to being carried out using SURF algorithm
Sign point extraction and matching are illustrated.
SURF algorithm is a kind of sane local feature region detection and description algorithm, the algorithm improvement extraction of feature and
Describing mode, the extraction and description of feature are completed with a kind of highly efficient mode, and specific implementation flow is as follows：
(1) Hessian (Hessian matrix) is built, generates all points of interest, the extraction for feature：
The purpose for building Hessian matrixes is in order to generate the marginal point of image stabilization (catastrophe point), with Canny, La Pu
The effect of Lars rim detection is similar, and basis is carried out for feature extraction hereafter.Hessian matrix (Hessian Matrix) is one
The square formation that the second-order partial differential coefficient of the individual function of many variables is formed, describe the local curvature of function.For an image f (x, y), its
Hessian matrixes are as follows：
When the discriminate of Hessian matrixes obtains local maximum, judge that current point is than other points in surrounding neighbors
Brighter or darker point, thus position the position of key point.It is above-mentioned usual before Hessian matrixes are constructed during specific implementation
Need to be filtered image, the mode of filtering can be selected according to actual calculated case, such as gaussian filtering, or boxlike
Filtering, builds corresponding filter operator and is calculated, specifically may be referred to related data of the prior art, the embodiment of the present invention
This is not limited.
(2) metric space is built：
The metric space of SURF algorithm is made up of O (Octaves, group) and L (Layer, layer), the size of image between different groups
All it is consistent, the difference is that, when being filtered using boxlike, the template size of the box filter used between different groups gradually increases
Greatly, different interlayers use the wave filter of identical size between same group, but the fuzzy coefficient of wave filter gradually increases.
(3) positioning feature point：
By by each pixel of Hessian matrix disposals and 26 in two dimensional image space and metric space neighborhood
Point is compared, and Primary Location goes out key point, then the filtered key point weak except energy comparison and the key point of location of mistake,
Filter out the characteristic point of final stabilization.
(4) characteristic point principal direction is distributed：
Using the haar wavelet characters in statistical nature point circle shaped neighborhood region, i.e., in the circle shaped neighborhood region of characteristic point, statistics 60
Horizontal, the vertical haar wavelet character summations of institute a little in degree is fan-shaped, then sector rotated with the interval of 0.2 radian size
And count again in the region after haar wavelet characters value, it will finally be worth that maximum fan-shaped direction as this feature point
Principal direction.
(5) feature point description is generated：
In SURF algorithm, 4*4 rectangular area block is taken around characteristic point, but acquired rectangular area direction is
Along the principal direction of characteristic point.The haar wavelet characters horizontally and vertically of 25 pixels are counted per sub-regions,
Here be all both horizontally and vertically relative principal direction for.The haar wavelet characters be horizontal direction value after, it is vertical
After direction value, after horizontal direction absolute value and 4 directions of vertical direction absolute value sum, using this 4 values as each
The characteristic vector in sub-block region, so description of the shared 4*4*4=64 dimensional vectors as SURF algorithm characteristic point.
(6) Feature Points Matching：
SURF algorithm is to determine matching degree by calculating the Euclidean distance between two characteristic points, and Euclidean distance is shorter, generation
The matching degree of two characteristic points of table is better.Further, SURF algorithm is also added into the judgement of Hessian traces of a matrix, if two
The trace of a matrix sign of characteristic point is identical, and representing the two features, there is the contrast on equidirectional to change, if it is different, saying
The contrast change direction of the two bright characteristic points is opposite, even if Euclidean distance is 0, is also directly excluded.
It should be appreciated that the above-mentioned description to SURF algorithm is only a kind of preferred form that the embodiment of the present invention is carried out, tool
When body is realized, associated materials of the prior art can also be referred to, carrying for local feature region is carried out by SURF algorithm to realize
Take and match, the embodiment of the present invention is not limited to this.
Further, it is contemplated that the quantity information of some target videos is more, and video is longer, and therefore, the above method also includes：
Target video is divided into multiple sub-goal videos；The frame of video of each sub-goal video is read frame by frame, with to each sub-goal
The key frame of video is judged.
Further, after the key frame of above-mentioned target video is judged, the above method also includes：Obtain the pass of target video
Key frame；The target video for including key frame is retrieved, to retrieve the video image similar to target video.
The decision method of a kind of key frame of video described in the embodiment of the present invention, by the video for reading target video frame by frame
Frame, it is reference frame to choose first frame of video, can extract the local feature region of reference frame, and obtain successively reference frame it is each after
Continuous frame, the mode that the local feature region of subsequent frame is matched with the local feature region of reference frame, can significantly be eliminated
Redundancy key frames, meanwhile, for the local feature region in different video stream, also can effectively be matched by conversion, and then sentence
Redundancy key frames are made, while redundancy key frames are significantly eliminated, calculating process is also simplify, improves video analysis
Efficiency.
Embodiment three：
On the basis of above-described embodiment, the embodiment of the present invention additionally provides a kind of decision maker of key frame of video, such as
A kind of structural representation of the decision maker of key frame of video shown in Fig. 3, the decision maker of the key frame of video include：With reference to
Frame flag module 31, First partial feature point extraction module 32, the second local feature point extraction module 33, matching judgment module 34
It is as follows with determination module 35, the function of modules：
Reference frame mark module 31, for reading the frame of video of target video frame by frame, by the first frame of video of target video
Labeled as reference frame；
First partial feature point extraction module 32, for extracting the local feature region of reference frame, and by local feature region mark
It is designated as First partial characteristic point；
Second local feature point extraction module 33, for obtaining each subsequent frame of reference frame successively, extraction is each follow-up
The local feature region of frame；
Matching judgment module 34, for judge the local feature region of each subsequent frame and First partial characteristic point whether
Match somebody with somebody；
Determination module 35, for when the judged result of matching judgment module is no, subsequent frame to be determined as into target video
Key frame.
Further, the structural representation of the decision maker of another key frame of video as shown in Figure 4, said apparatus also wrap
Include：
Reference frame chooses module 36, for after the first key frame of target video is determined, using key frame as reference frame,
Continue the key frame of judgement target video, until terminating when reference frame is without subsequent frame；
Module 37 is retrieved, for obtaining the key frame of target video；The target video for including key frame is retrieved,
To retrieve the video image similar to target video.
The decision maker of key frame of video provided in an embodiment of the present invention, the key frame of video provided with above-described embodiment
Decision method has identical technical characteristic, so can also solve identical technical problem, reaches identical technique effect.
Example IV：
The embodiment of the present invention additionally provides a kind of server, and the server includes processor, and memory, bus and communication connect
Mouthful, processor, communication interface and memory are connected by bus；Memory is used for storage program；Processor, for passing through bus
Call the program of storage in memory, the method described in execution above-described embodiment one or embodiment two.
Further, the embodiment of the present invention additionally provides a kind of computer-readable storage medium, for saving as above-mentioned Video Key
Computer software instructions used in the decision maker of frame, it includes the decision maker for being used for performing that above-mentioned aspect is key frame of video
Designed program.
Referring to Fig. 5, the embodiment of the present invention also provides a kind of structural representation of server, including：Processor 500, storage
Device 501, bus 502 and communication interface 503, processor 500, communication interface 503 and memory 501 are connected by bus 502；Place
Reason device 500 is used to perform the executable module stored in memory 501, such as computer program.
Wherein, memory 501 may include high-speed random access memory (RAM, Random Access Memory),
Non-labile memory (non-volatile memory), for example, at least a magnetic disk storage may also be included.By extremely
A few communication interface 503 (can be wired or wireless) is realized logical between the system network element and at least one other network element
Letter connection, can use internet, wide area network, LAN, Metropolitan Area Network (MAN) etc..
Bus 502 can be isa bus, pci bus or eisa bus etc..The bus can be divided into address bus, number
According to bus, controlling bus etc..For ease of representing, only represented in Fig. 5 with a four-headed arrow, it is not intended that an only bus
Or a type of bus.
Wherein, memory 501 is used for storage program, and processor 500 performs described program after execute instruction is received,
The method performed by device that the stream process that foregoing any embodiment of the embodiment of the present invention discloses defines can apply to processor
In 500, or realized by processor 500.
Processor 500 is probably a kind of IC chip, has the disposal ability of signal.It is above-mentioned in implementation process
Each step of method can be completed by the integrated logic circuit of the hardware in processor 500 or the instruction of software form.On
The processor 500 stated can be general processor, including central processing unit (Central Processing Unit, referred to as
CPU), network processing unit (Network Processor, abbreviation NP) etc.；It can also be digital signal processor (Digital
Signal Processing, abbreviation DSP), application specific integrated circuit (Application Specific Integrated
Circuit, abbreviation ASIC), ready-made programmable gate array (Field-Programmable Gate Array, abbreviation FPGA) or
Person other PLDs, discrete gate or transistor logic, discrete hardware components.It can realize or perform sheet
Disclosed each method, step and logic diagram in inventive embodiments.General processor can be microprocessor or the processing
Device can also be any conventional processor etc..The step of method with reference to disclosed in the embodiment of the present invention, can be embodied directly in
Hardware decoding processor performs completion, or performs completion with the hardware in decoding processor and software module combination.Software mould
Block can be located at random access memory, flash memory, read-only storage, programmable read only memory or electrically erasable programmable storage
In the ripe storage medium in this areas such as device, register.The storage medium is located at memory 501, and processor 500 reads memory
Information in 501, with reference to the step of its hardware completion above method.
The decision method for the key frame of video that the embodiment of the present invention is provided and the computer program product of device, including deposit
The computer-readable recording medium of program code is stored up, the instruction that described program code includes can be used for performing previous methods implementation
Method described in example, specific implementation can be found in embodiment of the method, will not be repeated here.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description
With the specific work process of device, the corresponding process in preceding method embodiment is may be referred to, will not be repeated here.
If the function is realized in the form of SFU software functional unit and is used as independent production marketing or in use, can be with
It is stored in a computer read/write memory medium.Based on such understanding, technical scheme is substantially in other words
The part to be contributed to prior art or the part of the technical scheme can be embodied in the form of software product, the meter
Calculation machine software product is stored in a storage medium, including some instructions are causing a computer equipment (can be
People's computer, server, or network equipment etc.) perform all or part of step of each embodiment methods described of the present invention.
And foregoing storage medium includes：USB flash disk, mobile hard disk, read-only storage (ROM, Read-Only Memory), arbitrary access are deposited
Reservoir (RAM, Random Access Memory), magnetic disc or CD etc. are various can be with the medium of store program codes.
Finally it should be noted that：Above example, it is only the embodiment of the present invention, to illustrate the skill of the present invention
Art scheme, rather than its limitations, protection scope of the present invention is not limited thereto, although entering with reference to the foregoing embodiments to the present invention
Go detailed description, it should be understood by those skilled in the art that：Any one skilled in the art takes off in the present invention
In the technical scope of dew, it can still modify to the technical scheme described in previous embodiment or can readily occur in change
Change, or equivalent substitution is carried out to which part technical characteristic；And these modifications, change or replacement, do not make relevant art
Scheme essence depart from technical scheme of the embodiment of the present invention spirit and scope, should all cover protection scope of the present invention it
It is interior.Therefore, protection scope of the present invention should be defined by scope of the claims.
Claims (10)
- A kind of 1. decision method of key frame of video, it is characterised in that including：The frame of video of target video is read frame by frame, is reference frame by the first video frame indicia of the target video；The local feature region of the reference frame is extracted, and the local feature region is labeled as First partial characteristic point；Each subsequent frame of the reference frame is obtained successively, extracts the local feature region of each subsequent frame；Judge whether the local feature region of each subsequent frame matches with the First partial characteristic point；If not, the subsequent frame is determined as to the key frame of the target video.
- 2. according to the method for claim 1, it is characterised in that methods described also includes：After the first key frame of the target video is determined, using the key frame as reference frame, continue described in judgement The key frame of target video, until terminating when the reference frame is without the subsequent frame.
- 3. according to the method for claim 1, it is characterised in that methods described includes：The local feature region is extracted using SURF algorithm, and judges the local feature region of each subsequent frame and described the Whether one local feature region matches.
- 4. according to the method for claim 3, it is characterised in that the local feature region for judging each subsequent frame with The First partial characteristic point whether match including：Judge that the local feature region of each subsequent frame is set in advance with whether the similarity of the First partial characteristic point is more than Fixed threshold value；If it is, determining that the subsequent frame matches with the reference frame, and the subsequent frame is labeled as redundancy key frames；If not, determine that the subsequent frame mismatches with the reference frame.
- 5. according to the method for claim 1, it is characterised in that methods described also includes：The target video is divided into multiple sub-goal videos；The frame of video of each sub-goal video is read frame by frame, to be carried out to the key frame of each sub-goal video Judge.
- 6. according to the method for claim 1, it is characterised in that methods described also includes：Obtain the key frame of the target video；The target video for including the key frame is retrieved, to retrieve the video figure similar to the target video Picture.
- A kind of 7. decision maker of key frame of video, it is characterised in that including：Reference frame mark module, for reading the frame of video of target video frame by frame, by the first video of the target video Frame flag is reference frame；First partial feature point extraction module, for extracting the local feature region of the reference frame, and by the local feature region Labeled as First partial characteristic point；Second local feature point extraction module, for obtaining each subsequent frame of the reference frame successively, extraction it is each it is described after The local feature region of continuous frame；Matching judgment module, for judge the local feature region of each subsequent frame and the First partial characteristic point whether Match somebody with somebody；Determination module, for when the judged result of the matching judgment module is no, the subsequent frame to be determined as into the mesh Mark the key frame of video.
- 8. device according to claim 7, it is characterised in that described device also includes：Reference frame chooses module, for after the first key frame of the target video is determined, using the key frame as Reference frame, continue to judge the key frame of the target video, until when the reference frame is without subsequent frame terminating.
- 9. device according to claim 7, it is characterised in that described device also includes：Module is retrieved, for obtaining the key frame of the target video；The target for including the key frame is regarded Frequency is retrieved, to retrieve the video image similar to the target video.
- 10. a kind of server, it is characterised in that the server includes processor, memory, bus and communication interface, described Processor, the communication interface and the memory are connected by the bus；The memory is used for storage program；The processor, for calling the program being stored in the memory by the bus, perform the claim Any methods describeds of 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710788529.8A CN107578011A (en) | 2017-09-05 | 2017-09-05 | The decision method and device of key frame of video |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710788529.8A CN107578011A (en) | 2017-09-05 | 2017-09-05 | The decision method and device of key frame of video |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107578011A true CN107578011A (en) | 2018-01-12 |
Family
ID=61029830
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710788529.8A Pending CN107578011A (en) | 2017-09-05 | 2017-09-05 | The decision method and device of key frame of video |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107578011A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110399842A (en) * | 2019-07-26 | 2019-11-01 | 北京奇艺世纪科技有限公司 | Method for processing video frequency, device, electronic equipment and computer readable storage medium |
| CN111046887A (en) * | 2018-10-15 | 2020-04-21 | 华北电力大学（保定） | Method for extracting characteristics of image with noise |
| CN111374607A (en) * | 2018-12-29 | 2020-07-07 | 尚科宁家(中国)科技有限公司 | Target identification method and device based on sweeping robot, equipment and medium |
| CN111491180A (en) * | 2020-06-24 | 2020-08-04 | 腾讯科技（深圳）有限公司 | Method and device for determining key frame |
| WO2020258499A1 (en) * | 2019-06-25 | 2020-12-30 | 平安科技（深圳）有限公司 | Object detection method and apparatus, and computer device and storage medium |
| CN114863692A (en) * | 2022-04-18 | 2022-08-05 | 广州天长信息技术有限公司 | Vehicle pattern recognition fee evasion checking method based on local feature alignment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070030396A1 (en) * | 2005-08-05 | 2007-02-08 | Hui Zhou | Method and apparatus for generating a panorama from a sequence of video frames |
| CN103761252A (en) * | 2013-12-25 | 2014-04-30 | 北京航天测控技术有限公司 | Video retrieval method |
| CN103810711A (en) * | 2014-03-03 | 2014-05-21 | 郑州日兴电子科技有限公司 | Keyframe extracting method and system for monitoring system videos |
| CN106203277A (en) * | 2016-06-28 | 2016-12-07 | 华南理工大学 | Fixed lens real-time monitor video feature extracting method based on SIFT feature cluster |
-
2017
- 2017-09-05 CN CN201710788529.8A patent/CN107578011A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070030396A1 (en) * | 2005-08-05 | 2007-02-08 | Hui Zhou | Method and apparatus for generating a panorama from a sequence of video frames |
| CN103761252A (en) * | 2013-12-25 | 2014-04-30 | 北京航天测控技术有限公司 | Video retrieval method |
| CN103810711A (en) * | 2014-03-03 | 2014-05-21 | 郑州日兴电子科技有限公司 | Keyframe extracting method and system for monitoring system videos |
| CN106203277A (en) * | 2016-06-28 | 2016-12-07 | 华南理工大学 | Fixed lens real-time monitor video feature extracting method based on SIFT feature cluster |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111046887A (en) * | 2018-10-15 | 2020-04-21 | 华北电力大学（保定） | Method for extracting characteristics of image with noise |
| CN111374607A (en) * | 2018-12-29 | 2020-07-07 | 尚科宁家(中国)科技有限公司 | Target identification method and device based on sweeping robot, equipment and medium |
| WO2020258499A1 (en) * | 2019-06-25 | 2020-12-30 | 平安科技（深圳）有限公司 | Object detection method and apparatus, and computer device and storage medium |
| CN110399842A (en) * | 2019-07-26 | 2019-11-01 | 北京奇艺世纪科技有限公司 | Method for processing video frequency, device, electronic equipment and computer readable storage medium |
| CN110399842B (en) * | 2019-07-26 | 2021-09-28 | 北京奇艺世纪科技有限公司 | Video processing method and device, electronic equipment and computer readable storage medium |
| CN111491180A (en) * | 2020-06-24 | 2020-08-04 | 腾讯科技（深圳）有限公司 | Method and device for determining key frame |
| CN114863692A (en) * | 2022-04-18 | 2022-08-05 | 广州天长信息技术有限公司 | Vehicle pattern recognition fee evasion checking method based on local feature alignment |
| CN114863692B (en) * | 2022-04-18 | 2023-10-27 | 广州天长信息技术有限公司 | Vehicle pattern recognition fee evasion checking method based on local feature alignment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yuan et al. | A robust and efficient approach to license plate detection | |
| CN107578011A (en) | The decision method and device of key frame of video | |
| CN110414507B (en) | License plate recognition method and device, computer equipment and storage medium | |
| Raghunandan et al. | Riesz fractional based model for enhancing license plate detection and recognition | |
| Gou et al. | Vehicle license plate recognition based on extremal regions and restricted Boltzmann machines | |
| US9014432B2 (en) | License plate character segmentation using likelihood maximization | |
| CN107665324B (en) | Image identification method and terminal | |
| WO2021051604A1 (en) | Method for identifying text region of osd, and device and storage medium | |
| Paunwala et al. | A novel multiple license plate extraction technique for complex background in Indian traffic conditions | |
| CN104408429A (en) | Method and device for extracting representative frame of video | |
| Bodnár et al. | Improving barcode detection with combination of simple detectors | |
| Zhang et al. | License plate localization in unconstrained scenes using a two-stage CNN-RNN | |
| CN108205657A (en) | Method, storage medium and the mobile terminal of video lens segmentation | |
| CN112417955A (en) | Patrol video stream processing method and device | |
| CN110969164A (en) | Low-illumination imaging license plate recognition method and device based on deep learning end-to-end | |
| CN114049499A (en) | Target object detection method, apparatus and storage medium for continuous contour | |
| CN111160107B (en) | Dynamic region detection method based on feature matching | |
| CN113435452A (en) | Electrical equipment nameplate text detection method based on improved CTPN algorithm | |
| Zhang et al. | Multi-scale segmentation strategies in PRNU-based image tampering localization | |
| CN108960246B (en) | Binarization processing device and method for image recognition | |
| CN111695373B (en) | Zebra stripes positioning method, system, medium and equipment | |
| CN114494990A (en) | Target detection method, system, terminal equipment and storage medium | |
| CN109978903A (en) | A kind of identification point recognition methods, device, electronic equipment and storage medium | |
| Khalid et al. | Image de-fencing using histograms of oriented gradients | |
| CN116580410A (en) | Bill number identification method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180112 |