CN106202130A - A kind of method and device of shot segmentation - Google Patents
A kind of method and device of shot segmentation Download PDFInfo
- Publication number
- CN106202130A CN106202130A CN201510234611.7A CN201510234611A CN106202130A CN 106202130 A CN106202130 A CN 106202130A CN 201510234611 A CN201510234611 A CN 201510234611A CN 106202130 A CN106202130 A CN 106202130A
- Authority
- CN
- China
- Prior art keywords
- camera lens
- adjacent picture
- picture
- segmentation
- characteristic point
- 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 discloses the method and device of a kind of shot segmentation.Wherein method includes: video carries out the primary segmentation of camera lens;Travel through every pair of adjacent picture during the adjacent shot transition after primary segmentation, for described every pair of adjacent picture, according to the matching degree of described adjacent picture, it is judged that whether described adjacent picture belongs to same camera lens;The most described adjacent picture is merged in same camera lens, otherwise, it determines described adjacent picture belongs to different camera lens.Instant invention overcomes the problem easily having influence on the accuracy of shot segmentation due to the angular transformation of object in same camera lens in prior art.
Description
Technical field
The present invention relates to multimedia technology field, particularly to the method and device of a kind of shot segmentation.
Background technology
The structure of video generally can be divided into four levels from high to low: video sequence, scene, camera lens, frame.
One video sequence generally refers to a single video file, or a video segment, video sequence
Being made up of several scenes, each scene comprises one or more camera lens, these camera lenses can be continuous print or
Person is spaced, and each camera lens includes some continuous print picture frames.Scene is in semantically phase from several
The video segment that can express certain semantic content that the camera lens closed collectively constitutes, these semantic relevant camera lenses
Can be continuous print, it is also possible to be interval.Camera lens is one section of video of video camera one-time continuous shooting, it
It it is the basic physics unit of video structural.Frame is the least unit of video, is the image of a width static state.Broadcast
When putting video file, fix picture at any time, is a two field picture.Camera lens refers to that video camera is continuous
One section of video of shooting, is made up of some frames.In the image change within camera lens, its reason is usually shooting
The motion of head and the motion of object, and the brightness flop etc. of light source.
Video capture device gathers the visual information of physical world to after on computer, forms original video.
These original videos are single camera lenses.Numerous single camera lenses through Video editing software optimization process,
After shearing, split, defining the video that user finally sees, the video after having processed is typically by many
Individual camera lens forms.
Shot boundary is the frame between adjacent camera lens and camera lens, crosses the knot that a border can be a upper camera lens
Bundle frame or the start frame of next camera lens, can also be to be inserted at the head and the tail of camera lens by video boundaries software equally
The a series of visual effect frames entered.
In order to video being carried out structuring process, need via the video after Video editing software editor also
Originally being multiple single camera lens, this process is referred to as shot segmentation (Shot Segment) or Video segmentation
(Video Segment).The main task of shot segmentation is to find out the location of shot boundary, i.e. one camera lens
Initial frame number and terminate frame number, and also want to determine the change type of camera lens.Propose boundary frame
Position or time point, can accordingly video be split.So, shot segmentation is typically also called camera lens
Border detection.
Existing shot segmentation technique includes: pixel difference method, histogram method, Block Matching Algorithm and limit
The automatic identifying methods such as edge rate method.In current block matching method, owing to the object in same camera lens becomes
Change such as affine transformation, perspective transform because have that the impact of unknown fresh information can be easy to by shot segmentation is automatic
Recognition methods is divided in different camera lenses, and so, in same camera lens, the angular transformation of object easily affects
Accuracy to shot segmentation.
Summary of the invention
In view of the above problems, it is proposed that the present invention is to provide one to overcome the problems referred to above or at least in part
Solve the method and device of a kind of shot segmentation of the problems referred to above.
The method of a kind of shot segmentation that the present invention provides, including:
Video is carried out the primary segmentation of camera lens;
Traversal primary segmentation after adjacent shot transition time every pair of adjacent picture, for described every pair adjacent
Picture, according to the matching degree of described adjacent picture, it is judged that whether described adjacent picture belongs to same camera lens;
The most described adjacent picture is merged in same camera lens, otherwise, it determines described adjacent picture
Belong to different camera lens.
Further, the described primary segmentation that video is carried out camera lens, specifically include:
Set the parameter of primary segmentation;Described parameter meets: after described primary segmentation, actually belong to same
The picture of individual camera lens can be assigned in identical or different camera lens, and actually belongs to the picture of different camera lens
Can not be segmented in identical camera lens;
According to the parameter of the primary segmentation of described setting, video is carried out the primary segmentation of camera lens.
Further, for described every pair of adjacent picture, according to the matching degree of described adjacent picture, it is judged that institute
State whether adjacent picture belongs to same camera lens, specifically include:
Described adjacent picture is carried out respectively feature point detection, extracts characteristic point;
Judge whether matching degree reaches the ratio set more than the characteristic point of the threshold value set;
The most then determine whether whether described adjacent picture exists effective matching area and be effectively matched region
Account for the ratio of whole picture size more than the threshold value set;The described region that is effectively matched is area intensive Du He district
In territory, the matching degree of characteristic point is satisfied by the region imposed a condition;
If, it is determined that described adjacent picture belongs to same camera lens, otherwise, is not belonging to same camera lens.
Further, described feature point detection that described adjacent picture is carried out respectively, after extracting characteristic point,
Also include:
The characteristic point extracted is carried out the consistent RANSAC of stochastic sampling purify.
Further, if judge the ratio that matching degree not up to sets more than the characteristic point of the threshold value set,
Determine that described adjacent picture is not belonging to same camera lens.
The device of a kind of shot segmentation that the present invention provides, including:
Primary segmentation module, for carrying out the primary segmentation of camera lens to video;
Judge module, every pair of adjacent picture during for traveling through the adjacent shot transition after primary segmentation, pin
To described every pair of adjacent picture, according to the matching degree of described adjacent picture, determine whether described adjacent picture belongs to
In same camera lens;
Determine module, for when judge module is judged as YES, described adjacent picture is merged into same mirror
In Tou, otherwise, it determines described adjacent picture belongs to different camera lens.
Further, above-mentioned primary segmentation module, specifically for setting the parameter of primary segmentation;Described parameter
Meet: after described primary segmentation, the picture actually belonging to same camera lens can be assigned to identical or different
Camera lens in, and the picture actually belonging to different camera lens can not be segmented in identical camera lens;Set according to described
The parameter of fixed primary segmentation, carries out the primary segmentation of camera lens to video.
Further, above-mentioned judge module, specifically for described adjacent picture being carried out respectively feature point detection,
Extract characteristic point;Judge whether matching degree reaches the ratio set more than the characteristic point of the threshold value set;If so,
Then determine whether whether described adjacent picture exists effective matching area and be effectively matched region and account for whole figure
The ratio of sheet size is more than the threshold value set;The described region that is effectively matched is feature in area intensive degree and region
The matching degree of point is satisfied by the region imposed a condition;
If, it is determined that described adjacent picture belongs to same camera lens, otherwise, is not belonging to same camera lens.
Further, above-mentioned judge module, it is additionally operable to carry out described adjacent picture respectively feature point detection,
After extracting characteristic point, the characteristic point extracted is carried out the consistent RANSAC of stochastic sampling and purifies.
Further, above-mentioned judge module, if being additionally operable to the characteristic point judging matching degree more than the threshold value set
During the ratio not up to set, determine that described adjacent picture is not belonging to same camera lens.
The beneficial effect comprise that
The method and device of the above-mentioned shot segmentation that the embodiment of the present invention provides, is first carried out at the beginning of camera lens video
Step segmentation;Then every pair of adjacent picture during the adjacent shot transition after primary segmentation is traveled through, for every pair
Adjacent picture, according to the matching degree of adjacent picture, it is judged that whether adjacent picture belongs to same camera lens;If so,
Then adjacent picture is merged in same camera lens, otherwise, it determines adjacent picture belongs to different camera lens.This
First inventive embodiments has carried out the shot segmentation of the first summary of fragmentation to video, then after recycling segmentation
The matching degree of the adjacent photo of shot transition, confirms whether both belong to same camera lens, if belonged to
Same camera lens, then correct the result of the shot segmentation omited at the beginning of early stage, both be grouped under same camera lens.
At the beginning of this elder generation, slightly fragmentation splits the method that the matching degree further according to the picture of shot transition carries out correcting, and carries
The high accuracy of shot segmentation, overcomes in prior art due to the angular transformation of object in same camera lens
The problem easily having influence on the accuracy of shot segmentation.
Other features and advantages of the present invention will illustrate in the following description, and, partly from explanation
Book becomes apparent, or understands by implementing the present invention.The purpose of the present invention and other advantages can
Realize by structure specifically noted in the description write, claims and accompanying drawing and obtain
?.
Below by drawings and Examples, technical scheme is described in further detail.
Accompanying drawing explanation
By reading the detailed description of hereafter preferred implementation, various other advantage and benefit for ability
Territory those of ordinary skill will be clear from understanding.Accompanying drawing is only used for illustrating the purpose of preferred implementation, and also
It is not considered as limitation of the present invention.And in whole accompanying drawing, it is denoted by the same reference numerals identical
Parts.In the accompanying drawings:
Fig. 1 is the flow chart of the method for shot segmentation in the embodiment of the present invention;
Fig. 2 is the flow chart of the primary segmentation that video carries out in the embodiment of the present invention camera lens;
Fig. 3 is the matching degree in the embodiment of the present invention according to described adjacent picture, it is judged that whether adjacent picture belongs to
Flow chart in same camera lens;
Fig. 4 is the flow chart that in the embodiment of the present invention, whether adjacent picture exists effective matching area;
Fig. 5 is the structural representation of the device of shot segmentation in the embodiment of the present invention;.
Detailed description of the invention
It is more fully described the exemplary embodiment of the disclosure below with reference to accompanying drawings.Although accompanying drawing shows
The exemplary embodiment of the disclosure, it being understood, however, that may be realized in various forms the disclosure and should be by
Embodiments set forth here is limited.On the contrary, it is provided that these embodiments are able to be best understood from this
Open, and complete for the scope of the present disclosure can be conveyed to those skilled in the art.
First the detailed description of the invention of the determination method of the lens type that the embodiment of the present invention provides is said
Bright.
The method of the shot segmentation that the embodiment of the present invention provides, as it is shown in figure 1, include:
S11, video is carried out the primary segmentation of camera lens;
S12, traversal primary segmentation after adjacent shot transition time every pair of adjacent picture, for every pair of phase
Adjacent picture, according to the matching degree of adjacent picture, it is judged that whether adjacent picture belongs to same camera lens;If so,
Perform following S13, if it is not, perform following S14;
S13, adjacent picture is merged in same camera lens;
S14, determine that adjacent picture belongs to different camera lens.
Separately below above steps is described in detail.
Further, above-mentioned S11 carries out to video the primary segmentation of camera lens, as in figure 2 it is shown, specifically wrap
Include:
S21, the parameter of setting primary segmentation;The parameter of primary segmentation need to meet: after primary segmentation, actual
On belong to the picture of same camera lens and can be assigned in identical or different camera lens, and actually belong to difference
The picture of camera lens can not be segmented in identical camera lens;
In the step s 21, can by use in prior art the most common in the way of carry out Shot change, arrange
Suitable parameters makes shot segmentation more fragmentation, the i.e. result of camera lens primary segmentation make the picture of a camera lens
Can be assigned in different camera lenses or a camera lens, but the picture of different camera lens can not be assigned to one
Rise.So so that follow-up same camera lens but become the picture of different adjacent camera lens to enter by primary segmentation
One step ground carries out refinement and corrects merging, until all camera lenses are all segmented correctly.
S22, according to the parameter of primary segmentation set, video is carried out the primary segmentation of camera lens.
Further, in above-mentioned S12, for described every pair of adjacent picture, according to described adjacent picture
Degree of joining, it is judged that whether described adjacent picture belongs to same camera lens, in the specific implementation, can pass through such as figure
Flow process shown in 3 realizes:
S31, described adjacent picture is carried out respectively feature point detection, extract characteristic point;
S32, judge whether matching degree reaches the ratio set more than the characteristic point of threshold value set;If so,
Then perform S33, otherwise, turn to execution S35;
If the characteristic point of the two adjacent picture coupling accounts for overall ratio very little, then it is assumed that both are not belonging to same
One camera lens, therefore, it can account for overall ratio less than setting in the characteristic point judging two adjacent picture couplings
Threshold value time, directly determine that this two pictures is not belonging to same camera lens.
S33, determine whether whether adjacent picture exists effective matching area and be effectively matched region and account for whole
The ratio of picture size is more than the threshold value set;Wherein, it is effectively matched in region is area intensive degree and region
The matching degree of characteristic point is satisfied by the region imposed a condition;The most then perform S35, otherwise, perform S36;
S34, determine that adjacent picture belongs to same camera lens;
S35, determine that adjacent picture is not belonging to same camera lens.
It is preferred that between above-mentioned S31 and S32, it is also possible to perform following step: the feature extracted is clicked on
Row stochastic sampling consistent (RANdom SAmple Consensus, RANSAC) purifies, RANSAC
The purpose purified is to remove the noise of the characteristic point extracted.
Specifically, in S33, determine whether whether adjacent picture exists effective matching area, Ke Yitong
Cross flow process as shown in Figure 4 to realize:
S41, respectively the setting regions in two two field pictures is carried out piecemeal, and according to the closeness of each piece, really
Close quarters in fixed described setting regions;
It is preferred that in above-mentioned S21, the setting regions in two two field pictures can be in before and after two frame picture, edge
The region accounting for whole picture 4/5 size of center picture radiation.Why use above-mentioned setting regions, be because of
For the data stream fps=25 of general video, data sampling rate is not less than 5 frames per second, so for cinestrip,
Do not worry the problem that adjacent picture change is the fastest.
S42, each characteristic point being respectively compared in the close quarters of two frame pictures, determine having in two frame pictures
Effect close quarters;Effectively in the area intensive degree in close quarters and region, the matching degree of characteristic point is satisfied by
Impose a condition;
S43, compare the block of same position in effective close quarters of two frame pictures, if in former frame picture
The characteristic point that block comprises, does not exists in the block of a later frame image same position, then two frame pictures before and after deleting
Middle characteristic of correspondence point.
Specifically, in the region of whole picture 4/5 size, carry out piecemeal further, such as, be divided into
Area1, Area2 ... etc., each block is arranged according to closeness, according to the threshold value of the closeness set,
Removing the block that closeness is relatively low, the region of remaining each piece of composition is exactly close quarters.
In close quarters, needing to further determine that effective close quarters, above-mentioned steps S42 is being embodied as
Time, the matching degree of all characteristic points in close quarters is arranged, the matching degree that same use is preset
Threshold value, screens out the characteristic point that some matching degrees are the highest, and remaining region has just obtained effective close quarters.
Based on same inventive concept, the embodiment of the present invention additionally provides the device of a kind of shot segmentation, due to this
The principle of the solved problem of device is similar to the method for aforementioned shot segmentation, and therefore the enforcement of this device can be joined
See the enforcement of preceding method, repeat no more in place of repetition.
The device of a kind of shot segmentation that the embodiment of the present invention provides, as it is shown in figure 5, include:
Primary segmentation module 501, for carrying out the primary segmentation of camera lens to video;
Judge module 502, every pair of adjacent picture during for traveling through the adjacent shot transition after primary segmentation,
For described every pair of adjacent picture, according to the matching degree of described adjacent picture, whether determine described adjacent picture
Belong to same camera lens;
Determine module 503, for when judge module is judged as YES, described adjacent picture is merged into same
In individual camera lens, otherwise, it determines described adjacent picture belongs to different camera lens.
Further, above-mentioned primary segmentation module 501, specifically for setting the parameter of primary segmentation;Described
Parameter meets: after described primary segmentation, actually belong to the picture of same camera lens can be assigned to identical or
In different camera lenses, and the picture actually belonging to different camera lens can not be segmented in identical camera lens;According to institute
State the parameter of the primary segmentation of setting, video is carried out the primary segmentation of camera lens.
Further, above-mentioned judge module 502, specifically for carrying out characteristic point respectively by described adjacent picture
Detection, extracts characteristic point;Judge whether matching degree reaches the ratio set more than the characteristic point of the threshold value set;
The most then determine whether whether described adjacent picture exists effective matching area and be effectively matched region and account for whole
The ratio of individual picture size is more than the threshold value set;Described it is effectively matched in region is area intensive degree and region
The matching degree of characteristic point is satisfied by the region imposed a condition;If, it is determined that described adjacent picture belongs to same
Individual camera lens, otherwise, is not belonging to same camera lens.
Further, above-mentioned judge module 502, it is additionally operable to carry out described adjacent picture respectively feature spot check
Survey, after extracting characteristic point, the characteristic point extracted is carried out the consistent RANSAC of stochastic sampling and purifies.
Further, above-mentioned judge module, if being additionally operable to the characteristic point judging matching degree more than the threshold value set
During the ratio not up to set, determine that described adjacent picture is not belonging to same camera lens.
The method and device of the above-mentioned shot segmentation that the embodiment of the present invention provides, is first carried out at the beginning of camera lens video
Step segmentation;Then every pair of adjacent picture during the adjacent shot transition after primary segmentation is traveled through, for every pair
Adjacent picture, according to the matching degree of adjacent picture, it is judged that whether adjacent picture belongs to same camera lens;If so,
Then adjacent picture is merged in same camera lens, otherwise, it determines adjacent picture belongs to different camera lens.This
First inventive embodiments has carried out the shot segmentation of the first summary of fragmentation to video, then after recycling segmentation
The matching degree of the adjacent photo of shot transition, confirms whether both belong to same camera lens, if belonged to
Same camera lens, then correct the result of the shot segmentation omited at the beginning of early stage, both be grouped under same camera lens.
At the beginning of this elder generation, slightly fragmentation splits the method that the matching degree further according to the picture of shot transition carries out correcting, and carries
The high accuracy of shot segmentation, overcomes in prior art due to the angular transformation of object in same camera lens
The problem easily having influence on the accuracy of shot segmentation.
Further, in the method and device of the above-mentioned shot segmentation of embodiment of the present invention offer, by judging phase
Whether adjacent picture exists effective matching area (is satisfied by for the matching degree of characteristic point in area intensive degree and region
The region imposed a condition) and be effectively matched region and account for the ratio of whole picture size and come really more than the threshold value set
The matching of fixed two adjacent photos, further improves the accuracy whether adjacent picture mates, enters
And ensure that the accuracy of shot segmentation.
Those skilled in the art are it should be appreciated that embodiments of the invention can be provided as method, system or meter
Calculation machine program product.Therefore, the present invention can use complete hardware embodiment, complete software implementation or knot
The form of the embodiment in terms of conjunction software and hardware.And, the present invention can use and wherein wrap one or more
Computer-usable storage medium containing computer usable program code (include but not limited to disk memory and
Optical memory etc.) form of the upper computer program implemented.
The present invention is with reference to method, equipment (system) and computer program product according to embodiments of the present invention
The flow chart of product and/or block diagram describe.It should be understood that can by computer program instructions flowchart and
/ or block diagram in each flow process and/or flow process in square frame and flow chart and/or block diagram and/
Or the combination of square frame.These computer program instructions can be provided to general purpose computer, special-purpose computer, embedding
The processor of formula datatron or other programmable data processing device is to produce a machine so that by calculating
The instruction that the processor of machine or other programmable data processing device performs produces for realizing at flow chart one
The device of the function specified in individual flow process or multiple flow process and/or one square frame of block diagram or multiple square frame.
These computer program instructions may be alternatively stored in and computer or the process of other programmable datas can be guided to set
In the standby computer-readable memory worked in a specific way so that be stored in this computer-readable memory
Instruction produce and include the manufacture of command device, this command device realizes in one flow process or multiple of flow chart
The function specified in flow process and/or one square frame of block diagram or multiple square frame.
These computer program instructions also can be loaded in computer or other programmable data processing device, makes
Sequence of operations step must be performed to produce computer implemented place on computer or other programmable devices
Reason, thus the instruction performed on computer or other programmable devices provides for realizing flow chart one
The step of the function specified in flow process or multiple flow process and/or one square frame of block diagram or multiple square frame.
Obviously, those skilled in the art can carry out various change and modification without deviating from this to the present invention
The spirit and scope of invention.So, if these amendments of the present invention and modification belong to the claims in the present invention
And within the scope of equivalent technologies, then the present invention is also intended to comprise these change and modification.
Claims (10)
1. the method for a shot segmentation, it is characterised in that including:
Video is carried out the primary segmentation of camera lens;
Traversal primary segmentation after adjacent shot transition time every pair of adjacent picture, for described every pair adjacent
Picture, according to the matching degree of described adjacent picture, it is judged that whether described adjacent picture belongs to same camera lens;
The most described adjacent picture is merged in same camera lens, otherwise, it determines described adjacent picture
Belong to different camera lens.
2. the method for claim 1, it is characterised in that described video is carried out the preliminary of camera lens
Segmentation, specifically includes:
Set the parameter of primary segmentation;Described parameter meets: after described primary segmentation, actually belong to same
The picture of individual camera lens can be assigned in identical or different camera lens, and actually belongs to the picture of different camera lens
Can not be segmented in identical camera lens;
According to the parameter of the primary segmentation of described setting, video is carried out the primary segmentation of camera lens.
3. method as claimed in claim 1 or 2, it is characterised in that for described every pair of adjacent picture,
Matching degree according to described adjacent picture, it is judged that whether described adjacent picture belongs to same camera lens, specifically wraps
Include:
Described adjacent picture is carried out respectively feature point detection, extracts characteristic point;
Judge whether matching degree reaches the ratio set more than the characteristic point of the threshold value set;
The most then determine whether whether described adjacent picture exists effective matching area and be effectively matched region
Account for the ratio of whole picture size more than the threshold value set;The described region that is effectively matched is area intensive Du He district
In territory, the matching degree of characteristic point is satisfied by the region imposed a condition;
If, it is determined that described adjacent picture belongs to same camera lens, otherwise, is not belonging to same camera lens.
4. method as claimed in claim 3, carries out described adjacent picture feature point detection respectively, carries
After taking characteristic point, also include:
The characteristic point extracted is carried out the consistent RANSAC of stochastic sampling purify.
5. method as claimed in claim 3, it is characterised in that matching degree is more than the threshold set if judging
During the ratio that the characteristic point of value not up to sets, determine that described adjacent picture is not belonging to same camera lens.
6. the device of a shot segmentation, it is characterised in that including:
Primary segmentation module, for carrying out the primary segmentation of camera lens to video;
Judge module, every pair of adjacent picture during for traveling through the adjacent shot transition after primary segmentation, pin
To described every pair of adjacent picture, according to the matching degree of described adjacent picture, determine whether described adjacent picture belongs to
In same camera lens;
Determine module, for when judge module is judged as YES, described adjacent picture is merged into same mirror
In Tou, otherwise, it determines described adjacent picture belongs to different camera lens.
7. device as claimed in claim 6, it is characterised in that primary segmentation module, specifically for setting
Determine the parameter of primary segmentation;Described parameter meets: after described primary segmentation, actually belong to same camera lens
Picture can be assigned in identical or different camera lens, and the picture actually belonging to different camera lens can not divide
Cut in identical camera lens;According to the parameter of the primary segmentation of described setting, video is carried out preliminary point of camera lens
Cut.
Device the most as claimed in claims 6 or 7, it is characterised in that described judge module, specifically uses
In described adjacent picture being carried out respectively feature point detection, extract characteristic point;Judge that matching degree is more than setting
Whether the characteristic point of threshold value reaches the ratio set;The most then determine whether whether described adjacent picture is deposited
It is being effectively matched region and is being effectively matched region and accounts for the ratio of whole picture size more than the threshold value set;Described
Being effectively matched region is that the matching degree of characteristic point is satisfied by the region imposed a condition in area intensive degree and region;
If, it is determined that described adjacent picture belongs to same camera lens, otherwise, is not belonging to same camera lens.
9. device as claimed in claim 8, it is characterised in that described judge module, is additionally operable to institute
State adjacent picture and carry out feature point detection respectively, after extracting characteristic point, the characteristic point extracted is carried out at random
The consistent RANSAC that samples purifies.
10. device as claimed in claim 8, it is characterised in that described judge module, if being additionally operable to
When judging the ratio that matching degree not up to sets more than the characteristic point of the threshold value set, determine described adjacent picture
It is not belonging to same camera lens.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510234611.7A CN106202130A (en) | 2015-05-08 | 2015-05-08 | A kind of method and device of shot segmentation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510234611.7A CN106202130A (en) | 2015-05-08 | 2015-05-08 | A kind of method and device of shot segmentation |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106202130A true CN106202130A (en) | 2016-12-07 |
Family
ID=57459909
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510234611.7A Pending CN106202130A (en) | 2015-05-08 | 2015-05-08 | A kind of method and device of shot segmentation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106202130A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108495171A (en) * | 2018-04-03 | 2018-09-04 | 优视科技有限公司 | Method for processing video frequency and its device, storage medium, electronic product |
CN109447022A (en) * | 2018-11-08 | 2019-03-08 | 北京奇艺世纪科技有限公司 | A kind of lens type recognition methods and device |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102800095A (en) * | 2012-07-17 | 2012-11-28 | 南京特雷多信息科技有限公司 | Lens boundary detection method |
CN103578094A (en) * | 2012-07-20 | 2014-02-12 | 清华大学 | Shot segmentation method |
CN104182974A (en) * | 2014-08-12 | 2014-12-03 | 大连理工大学 | A speeded up method of executing image matching based on feature points |
-
2015
- 2015-05-08 CN CN201510234611.7A patent/CN106202130A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102800095A (en) * | 2012-07-17 | 2012-11-28 | 南京特雷多信息科技有限公司 | Lens boundary detection method |
CN103578094A (en) * | 2012-07-20 | 2014-02-12 | 清华大学 | Shot segmentation method |
CN104182974A (en) * | 2014-08-12 | 2014-12-03 | 大连理工大学 | A speeded up method of executing image matching based on feature points |
Non-Patent Citations (3)
Title |
---|
熊伟等: "使用线性分割和序列合并的视频镜头分割方法", 《电子学报》 * |
缪源: "图像匹配算法的研究", 《万方学位论文库》 * |
赵璐璐等: "基于SURF和快速近似最近邻搜索的图像匹配算法", 《计算机应用研究》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108495171A (en) * | 2018-04-03 | 2018-09-04 | 优视科技有限公司 | Method for processing video frequency and its device, storage medium, electronic product |
CN109447022A (en) * | 2018-11-08 | 2019-03-08 | 北京奇艺世纪科技有限公司 | A kind of lens type recognition methods and device |
CN109447022B (en) * | 2018-11-08 | 2021-04-09 | 北京奇艺世纪科技有限公司 | Lens type identification method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10671855B2 (en) | Video object segmentation by reference-guided mask propagation | |
Caelles et al. | One-shot video object segmentation | |
WO2021179898A1 (en) | Action recognition method and apparatus, electronic device, and computer-readable storage medium | |
Ma et al. | Stage-wise salient object detection in 360 omnidirectional image via object-level semantical saliency ranking | |
CN111091091A (en) | Method, device and equipment for extracting target object re-identification features and storage medium | |
US9542735B2 (en) | Method and device to compose an image by eliminating one or more moving objects | |
US10600158B2 (en) | Method of video stabilization using background subtraction | |
WO2021139197A1 (en) | Image processing method and apparatus | |
US9213898B2 (en) | Object detection and extraction from image sequences | |
Voigtlaender et al. | Boltvos: Box-level tracking for video object segmentation | |
US20220309623A1 (en) | Method and apparatus for processing video | |
Voigtlaender et al. | Reducing the annotation effort for video object segmentation datasets | |
US20220083808A1 (en) | Method and apparatus for processing images, device and storage medium | |
CN111476234A (en) | Method and device for recognizing characters of shielded license plate, storage medium and intelligent equipment | |
CN106469293A (en) | The method and system of quick detection target | |
CN106202130A (en) | A kind of method and device of shot segmentation | |
EP3043315A1 (en) | Method and apparatus for generating superpixels for multi-view images | |
CN108369640A (en) | For control scene capture images image procossing to adjust the method, apparatus or computer program of capture images | |
CN111192286A (en) | Image synthesis method, electronic device and storage medium | |
Ghorpade et al. | Extracting text from video | |
KR20210067824A (en) | Method for single image dehazing based on deep learning, recording medium and device for performing the method | |
Jin et al. | Object-based video forgery detection via dual-stream networks | |
CN104182959A (en) | Target searching method and target searching device | |
CN106203244B (en) | A kind of determination method and device of lens type | |
CN113992979B (en) | Video expansion method and system and computer equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161207 |
|
RJ01 | Rejection of invention patent application after publication |