CN105791705B - Video anti-fluttering method, system and camera terminal suitable for mobile time-lapse photography - Google Patents
Video anti-fluttering method, system and camera terminal suitable for mobile time-lapse photography Download PDFInfo
- Publication number
- CN105791705B CN105791705B CN201610357076.9A CN201610357076A CN105791705B CN 105791705 B CN105791705 B CN 105791705B CN 201610357076 A CN201610357076 A CN 201610357076A CN 105791705 B CN105791705 B CN 105791705B
- Authority
- CN
- China
- Prior art keywords
- video
- frame
- homography matrix
- key
- key 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.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/68—Control of cameras or camera modules for stable pick-up of the scene, e.g. compensating for camera body vibrations
- H04N23/681—Motion detection
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/68—Control of cameras or camera modules for stable pick-up of the scene, e.g. compensating for camera body vibrations
- H04N23/681—Motion detection
- H04N23/6811—Motion detection based on the image signal
Abstract
The invention discloses a kind of video anti-fluttering methods suitable for mobile time-lapse photography, system and camera terminal, it carries out key point extraction by the picture frame to video to be processed and key point matches, the homography matrix between the adjacent image frame is calculated according to the matched result of key point, and the motion profile of the video to be processed is calculated according to the homography matrix of the adjacent image frame, then the key frame in the picture frame of video to be processed is extracted according to the motion profile, the homography matrix of the key frame is smoothed to obtain and stablizes video, it only utilizes video image information, it carries out taking out frame and smoothing processing in conjunction with the key point information and homography relationship of image, hardware is not depended on, algorithm robustness is high, anti-shake effect is significant, especially suitable for shaking biggish mobile time-lapse photography, and it also fits Stabilization for ordinary video is handled.
Description
Technical field
The present invention relates to technical field of video processing, especially a kind of video stabilization side suitable for mobile time-lapse photography
The system of method and its application this method, camera terminal.
Background technique
Mobile time-lapse photography is a branch of time-lapse photography, is a kind of fast-changing photography hand of performance photographic subjects
Method has extremely strong time compressed capability, prolonged things can be changed in compression in a short period of time and be displayed.With
Traditional time-lapse photography is different using the equipment such as tripod fixed camera, and mobile time-lapse photography is during shooting, camera
Equal capture apparatus are mostly man-hour manually hand-held, and track shot person constantly moves.Due to the unstability and compression view of hand-held photography
The process of frequency time will cause the irregular high dither phenomenon of photographic.In order to improve the shooting effect of mobile time-lapse photography
Fruit needs to carry out stabilization processing to shooting video.
In the video anti-fluttering method of the prior art, generally by the gyro data and video figure for obtaining mobile device
As data, then video image data is compensated in turn according to the motion conditions of gyro data estimation mobile device, is cut out
The processing such as cut.This stabilization mode based on gyroscope is to the more demanding of hardware, situations such as being easy to appear zero and float, be delayed, and
It is undesirable for the anti-shake effect significantly shaken.
Summary of the invention
The present invention is to solve the above problems, providing a kind of video anti-fluttering method suitable for mobile time-lapse photography, being
System and camera terminal mainly utilize video image information, do not depend on hardware, algorithm robustness is higher, for different scenes
Video image stabilization processing all have preferable effect.
To achieve the above object, the technical solution adopted by the present invention are as follows:
Firstly, the present invention provides a kind of video anti-fluttering method suitable for mobile time-lapse photography comprising following steps:
10. the picture frame of pair video to be processed carries out key point extraction;
20. carrying out key point matching to adjacent image frame according to extracted key point;
30. matched according to the key point as a result, calculating the homography matrix between the adjacent image frame;
40. calculating the motion profile of the video to be processed according to the homography matrix of the adjacent image frame;
50. extracting the key frame in the picture frame of video to be processed according to the motion profile;
60. the homography matrix of pair key frame is smoothed, obtain stablizing video.
Preferably, in the step 10, key point extraction is carried out to the picture frame of video to be processed, be by traversal to
Each frame picture frame of video is handled, and each pixel and neighbor pixel of described image frame are carried out to the ratio of gray value
Compared with, and according to the label of the gray scale difference value of each pixel and neighbor pixel progress key point.
Preferably, in the step 20, key point matching is carried out to adjacent image frame, further comprises:
21. calculating the histogram of gradients of current key neighborhood of a point range, and using the histogram of gradients as current key
The descriptor of point;
22. carrying out of similitude to the correspondence key point of the adjacent image frame according to the descriptor of the key point
Match.
Preferably, in the step 30, homography matrix between the adjacent image frame is by described adjacent
The correspondence key point of picture frame to match carries out the calculating of linear transformation, obtains the homography square between the adjacent image frame
Battle array, calculation formula are as follows:
X ,=Hx;
Wherein, x, is the set of keypoints of picture frame to be matched, and x is the corresponding set of keypoints of reference image frame, H
For the homography matrix.
Preferably, in the step 40, the view to be processed is calculated according to the homography matrix of the adjacent image frame
The motion profile of frequency is to handle to obtain institute by the way that the homography matrix of all picture frames before current image frame is carried out multiplication
The motion state of current image frame is stated, so that the motion state of each picture frame of the video to be processed is calculated, finally
The motion state of each picture frame is fitted to obtain the motion profile of the video to be processed.
Preferably, in the step 50, the pass in the picture frame of video to be processed is extracted according to the motion profile
Key frame calculates the spacing distance of key frame to be extracted yet further still according to the acceleration degree of time-lapse photography, according to the movement
Track and the spacing distance carry out the extraction of key frame.
Preferably, in the step 60, the calculation formula of the homography matrix of the key frame is as follows:
Wherein, h indicates the homography matrix of three rows three column, and the index number 1,2,3 of h indicates the homography matrix
The 1st, 2,3 rows, T indicates that the transposition of the homography matrix, x, y indicate the respective pixel before the affine transformation of the key frame
The coordinate value of point, the coordinate value of the corresponding pixel points after the affine transformation of x ', y ' the expression key frame.
Preferably, in the step 60, the homography matrix of the key frame is smoothed, is by that will put down
Sliding treated that homography matrix obtains amendment transformation matrix divided by the homography matrix before smoothing processing, and uses the amendment
Transformation matrix carries out interpolation calculation to the key frame, obtains stablizing video.
Secondly, the present invention provides a kind of video stabilization system suitable for mobile time-lapse photography comprising:
Key point extraction module carries out key point extraction for the picture frame to video to be processed;
Key point matching module carries out key point matching to adjacent image frame according to extracted key point;
Homography matrix solves module, matched according to the key point as a result, calculating the adjacent image frame
Between homography matrix;
Motion profile solves module, according to the homography matrix of the adjacent image frame, calculates the video to be processed
Motion profile;
Key Frame Extraction module extracts the key frame in the picture frame of video to be processed according to the motion profile;
Smoothing module is smoothed for the homography matrix to the key frame, obtains stablizing video.
In addition, the present invention also provides a kind of camera terminals comprising be suitable for mobile time-lapse photography as described above
Video stabilization system.
The beneficial effects of the present invention are:
A kind of video anti-fluttering method, system and camera terminal suitable for mobile time-lapse photography of the invention, passes through
Key point extraction and key point matching are carried out to the picture frame of video to be processed, calculated according to the result of the Feature Points Matching
Homography matrix between the adjacent image frame out, and it is described wait locate according to the calculating of the homography matrix of the adjacent image frame
The motion profile for managing video, then extracts the key frame in the picture frame of video to be processed, to institute according to the motion profile
The homography matrix for stating key frame, which is smoothed to obtain, stablizes video, video image information is only utilized, in conjunction with image
Key point information and homography relationship carry out taking out frame and smoothing processing, do not depend on hardware, and algorithm robustness is high, and anti-shake effect is aobvious
It writes, especially suitable for shaking biggish mobile time-lapse photography, and is also applied for the stabilization processing of ordinary video.
Detailed description of the invention
The drawings described herein are used to provide a further understanding of the present invention, constitutes a part of the invention, this hair
Bright illustrative embodiments and their description are used to explain the present invention, and are not constituted improper limitations of the present invention.In the accompanying drawings:
Fig. 1 is the general flow chart for the video anti-fluttering method that the present invention is suitable for mobile time-lapse photography;
Fig. 2 is the structural schematic diagram for the video stabilization system that the present invention is suitable for mobile time-lapse photography.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.It should be appreciated that specific embodiment described herein is only to solve
The present invention is released, is not intended to limit the present invention.Based on the embodiments of the present invention, those of ordinary skill in the art are not making
Every other embodiment obtained, shall fall within the protection scope of the present invention under the premise of creative work.
As shown in Figure 1, a kind of video anti-fluttering method suitable for mobile time-lapse photography of the invention comprising following step
It is rapid:
10. the picture frame of pair video to be processed carries out key point extraction;
20. carrying out key point matching to adjacent image frame according to extracted key point;
30. matched according to the key point as a result, calculating the homography matrix between the adjacent image frame;
40. calculating the motion profile of the video to be processed according to the homography matrix of the adjacent image frame;
50. extracting the key frame in the picture frame of video to be processed according to the motion profile;
60. the homography matrix of pair key frame is smoothed, obtain stablizing video.
In the step 10, key point extraction is carried out to the picture frame of video to be processed, is by traversing view to be processed
Each frame picture frame of frequency, and by each pixel of described image frame compared with neighbor pixel carries out gray value, and root
The label of key point is carried out according to the gray scale difference value of each pixel and neighbor pixel.The extracting method of image key points can be with
Using the other methods of the prior art, such as SIFT critical point detection can also be used, by using laplacian image gold
Word tower and gaussian filtering difference image quickly to seek the extreme value in Laplacian space.
In the step 20, key point matching is carried out to adjacent image frame, mainly according to the correspondence of adjacent image frame
The similitude of key point is matched, in the present embodiment, using following steps:
21. calculating the histogram of gradients of current key neighborhood of a point range, and using the histogram of gradients as current key
The descriptor of point;
22. carrying out of similitude to the correspondence key point of the adjacent image frame according to the descriptor of the key point
Match.
In the step 30, homography matrix between the adjacent image frame is by the adjacent image frame
The correspondence key point that matches carry out the calculating of linear transformation, obtain the homography matrix between the adjacent image frame, count
It is as follows to calculate formula:
X ,=Hx;
Wherein, x, is the set of keypoints of picture frame to be matched, and x is the corresponding set of keypoints of reference image frame, H
It is here by converting the coordinate of the correspondence key point of picture frame and reference image frame to be matched for the homography matrix
Secondary coordinate system, and the homography relationship between corresponding key point is calculated using least square method.
Alternatively, the present invention can also obtain the homography between the adjacent image frame using the calculating of nonlinear transformation
Matrix, such as by cutting image, it is assumed that it is cut into the fritter of 10 rows 10 column, one is calculated to each fritter
Homography matrix can make the matching effect of image finer in this way, can achieve better anti-shake effect, but operand is very
Greatly;Alternatively, can also be weighted by different types of homography matrix, such as affine transformation matrix with similitude transformation matrix
Summation can be anti-shake effect more robust, be not in big pattern distortion as final homography matrix.
In the step 40, the movement of the video to be processed is calculated according to the homography matrix of the adjacent image frame
Track is to handle to obtain the current figure by the way that the homography matrix of all picture frames before current image frame is carried out multiplication
As the motion state of frame, so that the motion state of each picture frame of the video to be processed is calculated, finally by each figure
As the motion state of frame is fitted to obtain the motion profile of the video to be processed.Homography matrix is indicated between consecutive frame
Variation relation, i.e. some translations, rotation, the variation of scaling of the next frame relative to previous frame, by the fortune for adding up each frame
Dynamic variation, to fit complete motion profile.The process that homography matrix calculates motion profile is exactly constantly to tire out to multiply singly to answer
Property matrix just pass through tired 1 Dao the n a homography matrix for multiplying front for example, to calculate n-th frame image state in which.
In the step 50, the key frame in the picture frame of video to be processed is extracted according to the motion profile, also
The spacing distance that key frame to be extracted is further calculated according to the acceleration degree of time-lapse photography, according to the motion profile and institute
State the extraction that spacing distance carries out key frame.Common time-lapse photography video is to carry out pumping frame according to fixed spacing distance, than
If setting and accelerating 2 times, it is exactly to extract an image every a frame exactly to take out a frame for video, loses a frame, take out one
Frame, as soon as to frame, the video ultimately produced forms the effect put fastly, and the pumping frame method of this simple violence in visual effect
It will lead to the huge shake between consecutive frame.The present invention carries out the pumping of key frame according to the motion profile and the spacing distance
It takes, stable video effect can be obtained.
In the step 60, the calculation formula of the homography matrix of the key frame is as follows:
Wherein, h indicates the homography matrix of three rows three column, and the index number 1,2,3 of h indicates the homography matrix
The 1st, 2,3 rows, T indicates that the transposition of the homography matrix, x, y indicate the respective pixel before the affine transformation of the key frame
The coordinate value of point, the coordinate value of the corresponding pixel points after the affine transformation of x ', y ' the expression key frame.
In the step 60, the homography matrix of the key frame is smoothed, is by by smoothing processing
Homography matrix afterwards obtains amendment transformation matrix divided by the homography matrix before smoothing processing, and converts square using the amendment
Battle array carries out interpolation calculation to the key frame, obtains stablizing video.
As shown in Fig. 2, the present invention also provides a kind of video stabilization system 100 suitable for mobile time-lapse photography, packet
It includes:
Key point extraction module 101 carries out key point extraction for the picture frame to video to be processed;
Key point matching module 102 carries out key point matching to adjacent image frame according to extracted key point;
Homography matrix solves module 103, matched according to the key point as a result, calculating the adjacent image
Homography matrix between frame;
Motion profile solves module 104, according to the homography matrix of the adjacent image frame, calculates the view to be processed
The motion profile of frequency;
Key Frame Extraction module 105 extracts the key in the picture frame of video to be processed according to the motion profile
Frame;
Smoothing module 106 is smoothed for the homography matrix to the key frame, obtains stablizing view
Frequently.
In addition, the camera terminal includes prolonging as described above suitable for movable type the present invention also provides a kind of camera terminal
When the video stabilization system photographed, wherein the video stabilization system suitable for mobile time-lapse photography can use Fig. 2 embodiment
Structure accordingly can execute the technical solution of embodiment of the method shown in Fig. 1, it is similar that the realization principle and technical effect are similar,
It may refer to the related record in above-described embodiment in detail, details are not described herein again.
The camera terminal includes: the equipment that mobile phone, digital camera or tablet computer etc. are configured with camera.
It should be noted that all the embodiments in this specification are described in a progressive manner, each embodiment weight
Point explanation is the difference from other embodiments, and the same or similar parts between the embodiments can be referred to each other.
For system embodiment and terminal embodiment, since it is basically similar to the method embodiment, so be described relatively simple,
The relevent part can refer to the partial explaination of embodiments of method.Also, herein, the terms "include", "comprise" or its
What his variant is intended to non-exclusive inclusion, so that including the process, methods of a series of elements, article or setting
Standby includes not only those elements, but also including other elements that are not explicitly listed, or further includes for this process, side
Method, article or the intrinsic element of equipment.In the absence of more restrictions, limited by sentence "including a ..."
Element, it is not excluded that there is also other identical elements in the process, method, article or apparatus that includes the element.Separately
Outside, those of ordinary skill in the art will appreciate that realizing that all or part of the steps of above-described embodiment can be by hardware come complete
At relevant hardware can also being instructed to complete by program, the program can store in a kind of computer-readable storage
In medium, storage medium mentioned above can be read-only memory, disk or CD etc..
The preferred embodiment of the present invention has shown and described in above description, it should be understood that the present invention is not limited to this paper institute
The form of disclosure, should not be regarded as an exclusion of other examples, and can be used for other combinations, modifications, and environments, and energy
Enough in this paper invented the scope of the idea, modifications can be made through the above teachings or related fields of technology or knowledge.And people from this field
The modifications and changes that member is carried out do not depart from the spirit and scope of the present invention, then all should be in the protection of appended claims of the present invention
In range.
Claims (9)
1. a kind of video anti-fluttering method suitable for mobile time-lapse photography, which comprises the following steps:
10. the picture frame of pair video to be processed carries out key point extraction;
20. carrying out key point matching to adjacent image frame according to extracted key point;
30. matched according to the key point as a result, calculating the homography matrix between the adjacent image frame;
40. calculating the motion profile of the video to be processed according to the homography matrix of the adjacent image frame;
50. extracting the key frame in the picture frame of video to be processed according to the motion profile, taken the photograph yet further still according to delay
The acceleration degree of shadow calculates the spacing distance of key frame to be extracted, is closed according to the motion profile and the spacing distance
The extraction of key frame;
60. the homography matrix of pair key frame is smoothed, obtain stablizing video.
2. a kind of video anti-fluttering method suitable for mobile time-lapse photography according to claim 1, it is characterised in that: institute
In the step 10 stated, key point extraction is carried out to the picture frame of video to be processed, is each frame by traversing video to be processed
Picture frame, and by each pixel of described image frame compared with neighbor pixel carries out gray value, and according to each pixel
Point carries out the label of key point with the gray scale difference value of neighbor pixel.
3. a kind of video anti-fluttering method suitable for mobile time-lapse photography according to claim 1, it is characterised in that: institute
In the step 20 stated, key point matching is carried out to adjacent image frame, further comprises:
21. calculating the histogram of gradients of current key neighborhood of a point range, and using the histogram of gradients as current key point
Descriptor;
22. carrying out the matching of similitude to the correspondence key point of the adjacent image frame according to the descriptor of the key point.
4. a kind of video anti-fluttering method suitable for mobile time-lapse photography according to claim 1, it is characterised in that: institute
In the step 30 stated, homography matrix between the adjacent image frame is by matching to the adjacent image frame
Corresponding key point carries out the calculating of linear transformation, obtains the homography matrix between the adjacent image frame, calculation formula is as follows:
X '=Hx;
Wherein, x ' is the set of keypoints of picture frame to be matched, and x is the corresponding set of keypoints of reference image frame, and H is institute
The homography matrix stated.
5. a kind of video anti-fluttering method suitable for mobile time-lapse photography according to claim 1, it is characterised in that: institute
In the step 40 stated, the motion profile of the video to be processed is calculated according to the homography matrix of the adjacent image frame, is logical
It crosses and the homography matrix of all picture frames before current image frame is subjected to multiplication handles to obtain the fortune of the current image frame
Dynamic state, so that the motion state of each picture frame of the video to be processed is calculated, finally by the fortune of each picture frame
Dynamic state is fitted to obtain the motion profile of the video to be processed.
6. a kind of video anti-fluttering method suitable for mobile time-lapse photography according to any one of claims 1 to 5, special
Sign is: in the step 60, the calculation formula of the homography matrix of the key frame is as follows:
Wherein, h indicates the homography matrix of three rows three column, and the index number 1,2,3 of h indicates the of the homography matrix
1,2,3 row, T indicate that the transposition of the homography matrix, x, y indicate the corresponding pixel points before the affine transformation of the key frame
Coordinate value, the coordinate value of the corresponding pixel points after the affine transformation of x ', y ' the expression key frame.
7. a kind of video anti-fluttering method suitable for mobile time-lapse photography according to any one of claims 1 to 5, special
Sign is: in the step 60, being smoothed to the homography matrix of the key frame, is by will be after smoothing processing
Homography matrix obtain amendment transformation matrix divided by the homography matrix before smoothing processing, and use the amendment transformation matrix
Interpolation calculation is carried out to the key frame, obtains stablizing video.
8. a kind of video stabilization system suitable for mobile time-lapse photography characterized by comprising
Key point extraction module carries out key point extraction for the picture frame to video to be processed;
Key point matching module carries out key point matching to adjacent image frame according to extracted key point;
Homography matrix solves module, matched according to the key point as a result, calculating between the adjacent image frame
Homography matrix;
Motion profile solves module, according to the homography matrix of the adjacent image frame, calculates the fortune of the video to be processed
Dynamic rail mark;
Key Frame Extraction module extracts the key frame in the picture frame of video to be processed according to the motion profile, also into
One step calculates the spacing distance of key frame to be extracted according to the acceleration degree of time-lapse photography, according to the motion profile and described
The extraction of spacing distance progress key frame;
Smoothing module is smoothed for the homography matrix to the key frame, obtains stablizing video.
9. a kind of camera terminal, which is characterized in that anti-including the video according to any one of claims 8 for being suitable for mobile time-lapse photography
Tremble system.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610357076.9A CN105791705B (en) | 2016-05-26 | 2016-05-26 | Video anti-fluttering method, system and camera terminal suitable for mobile time-lapse photography |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610357076.9A CN105791705B (en) | 2016-05-26 | 2016-05-26 | Video anti-fluttering method, system and camera terminal suitable for mobile time-lapse photography |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105791705A CN105791705A (en) | 2016-07-20 |
CN105791705B true CN105791705B (en) | 2019-06-11 |
Family
ID=56379515
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610357076.9A Active CN105791705B (en) | 2016-05-26 | 2016-05-26 | Video anti-fluttering method, system and camera terminal suitable for mobile time-lapse photography |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105791705B (en) |
Families Citing this family (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20190387166A1 (en) * | 2016-12-20 | 2019-12-19 | The University Of Tokyo | Image processing apparatus and program |
CN106961551A (en) * | 2017-03-23 | 2017-07-18 | 余仁集 | A kind of VR panoramic videos later stage stabilization technology |
CN107330410B (en) * | 2017-07-03 | 2020-06-30 | 南京工程学院 | Anomaly detection method based on deep learning in complex environment |
CN107809590B (en) * | 2017-11-08 | 2020-04-28 | 青岛海信移动通信技术股份有限公司 | Photographing method and device |
CN108600622B (en) * | 2018-04-12 | 2021-12-24 | 联想(北京)有限公司 | Video anti-shake method and device |
WO2020019212A1 (en) * | 2018-07-25 | 2020-01-30 | 深圳市大疆创新科技有限公司 | Video playback speed control method and system, control terminal, and mobile platform |
CN109089015B (en) * | 2018-09-19 | 2020-12-22 | 厦门美图之家科技有限公司 | Video anti-shake display method and device |
CN112396639A (en) * | 2019-08-19 | 2021-02-23 | 虹软科技股份有限公司 | Image alignment method |
CN110572534A (en) * | 2019-09-19 | 2019-12-13 | 浙江大搜车软件技术有限公司 | Digital video image stabilization method, device, equipment and storage medium of panoramic image |
CN115209031A (en) * | 2021-04-08 | 2022-10-18 | 北京字跳网络技术有限公司 | Video anti-shake processing method and device, electronic equipment and storage medium |
WO2022222113A1 (en) * | 2021-04-22 | 2022-10-27 | 深圳市大疆创新科技有限公司 | Video processing method, apparatus and system, and storage medium |
CN113612923B (en) * | 2021-07-30 | 2023-02-03 | 重庆电子工程职业学院 | Dynamic visual effect enhancement system and control method |
CN113689458B (en) * | 2021-10-27 | 2022-03-29 | 广州市玄武无线科技股份有限公司 | 2D shooting track path calculation method and device |
CN114095659B (en) * | 2021-11-29 | 2024-01-23 | 厦门美图之家科技有限公司 | Video anti-shake method, device, equipment and storage medium |
CN114257748A (en) * | 2022-01-26 | 2022-03-29 | Oppo广东移动通信有限公司 | Video anti-shake method and device, computer readable medium and electronic device |
CN114429191B (en) * | 2022-04-02 | 2023-03-28 | 深圳深知未来智能有限公司 | Electronic anti-shake method, system and storage medium based on deep learning |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101521740A (en) * | 2009-04-01 | 2009-09-02 | 北京航空航天大学 | Real-time athletic estimating method based on multiple dimensioned unchanged characteristic |
CN101854465A (en) * | 2010-02-01 | 2010-10-06 | 杭州海康威视软件有限公司 | Image processing method and device based on optical flow algorithm |
WO2013109335A1 (en) * | 2012-01-16 | 2013-07-25 | Google Inc. | Methods and systems for processing a video for stablization using dynamic crop |
CN103700069A (en) * | 2013-12-11 | 2014-04-02 | 武汉工程大学 | ORB (object request broker) operator-based reference-free video smoothness evaluation method |
CN105141807A (en) * | 2015-09-23 | 2015-12-09 | 北京二郎神科技有限公司 | Video signal image processing method and device |
CN105141872A (en) * | 2015-08-20 | 2015-12-09 | 成都鹰眼视觉科技有限公司 | Video image time-lapse processing method |
-
2016
- 2016-05-26 CN CN201610357076.9A patent/CN105791705B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101521740A (en) * | 2009-04-01 | 2009-09-02 | 北京航空航天大学 | Real-time athletic estimating method based on multiple dimensioned unchanged characteristic |
CN101854465A (en) * | 2010-02-01 | 2010-10-06 | 杭州海康威视软件有限公司 | Image processing method and device based on optical flow algorithm |
WO2013109335A1 (en) * | 2012-01-16 | 2013-07-25 | Google Inc. | Methods and systems for processing a video for stablization using dynamic crop |
CN103700069A (en) * | 2013-12-11 | 2014-04-02 | 武汉工程大学 | ORB (object request broker) operator-based reference-free video smoothness evaluation method |
CN105141872A (en) * | 2015-08-20 | 2015-12-09 | 成都鹰眼视觉科技有限公司 | Video image time-lapse processing method |
CN105141807A (en) * | 2015-09-23 | 2015-12-09 | 北京二郎神科技有限公司 | Video signal image processing method and device |
Non-Patent Citations (2)
Title |
---|
Re-Cinematography:Improving the Camerawork of Casual Video;MICHAEL L. GLEICHER 等;《ACM Transactions on Multimedia Computing, Communications, and Applications》;20081031;第5卷(第1期);全文 |
高效视频稳定与内容浏览方法研究;徐千昆;《中国优秀硕士学位论文全文数据库 信息科技辑》;20160315(第3期);全文 |
Also Published As
Publication number | Publication date |
---|---|
CN105791705A (en) | 2016-07-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105791705B (en) | Video anti-fluttering method, system and camera terminal suitable for mobile time-lapse photography | |
US20210377460A1 (en) | Automatic composition of composite images or videos from frames captured with moving camera | |
Yang et al. | Fast single image reflection suppression via convex optimization | |
Rengarajan et al. | Unrolling the shutter: Cnn to correct motion distortions | |
Elgharib et al. | Video magnification in presence of large motions | |
CN110827200A (en) | Image super-resolution reconstruction method, image super-resolution reconstruction device and mobile terminal | |
WO2020211624A1 (en) | Object tracking method, tracking processing method, corresponding apparatus and electronic device | |
CN105427333B (en) | Real-time Registration, system and the camera terminal of video sequence image | |
CN105635588B (en) | A kind of digital image stabilization method and device | |
CN107346414B (en) | Pedestrian attribute identification method and device | |
Lu et al. | Robust blur kernel estimation for license plate images from fast moving vehicles | |
CN112995678B (en) | Video motion compensation method and device and computer equipment | |
CN104537381B (en) | A kind of fuzzy image recognition method based on fuzzy invariant features | |
CN110248048A (en) | A kind of detection method and device of video jitter | |
Jeon et al. | Multi-image deblurring using complementary sets of fluttering patterns | |
CN111951325A (en) | Pose tracking method, pose tracking device and electronic equipment | |
CN107742278B (en) | Binding of L0Motion blurred image blind restoration method based on norm and spatial scale information | |
CN108229281B (en) | Neural network generation method, face detection device and electronic equipment | |
Bhattacharya et al. | Dehazing of color image using stochastic enhancement | |
CN110309721A (en) | Method for processing video frequency, terminal and storage medium | |
Kim et al. | Light field angular super-resolution using convolutional neural network with residual network | |
CN113888425A (en) | Industrial quality inspection image character matching method and system based on multi-frame fusion | |
CN111695525B (en) | 360-degree clothing fitting display method and device | |
Lau et al. | Atdetect: Face detection and keypoint extraction at range and altitude | |
Zhang et al. | Single Image Reflection Removal Based on Dark Channel Sparsity Prior |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |