CN104123396A - Soccer video abstract generation method and device based on cloud television - Google Patents
Soccer video abstract generation method and device based on cloud television Download PDFInfo
- Publication number
- CN104123396A CN104123396A CN201410401744.4A CN201410401744A CN104123396A CN 104123396 A CN104123396 A CN 104123396A CN 201410401744 A CN201410401744 A CN 201410401744A CN 104123396 A CN104123396 A CN 104123396A
- Authority
- CN
- China
- Prior art keywords
- video
- mass
- tone
- court
- audio frequency
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/73—Querying
- G06F16/738—Presentation of query results
- G06F16/739—Presentation of query results in form of a video summary, e.g. the video summary being a video sequence, a composite still image or having synthesized frames
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Image Analysis (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a soccer video abstract generation method and device based on cloud television. The method includes the steps that the excellence degrees of soccer videos are analyzed in real time, excellent video clips are determined and uploaded to the cloud end, and video abstracts are generated. Through the soccer video abstract generation method and device, real-time video abstracts can be combined with the Cloud PVR technology, and stress on the network and the cloud end is relieved.
Description
Technical field
The present invention relates to cloud TV technology, relate in particular to a kind of abstract of football video generation method and device based on cloud TV.
Background technology
Along with market is to possessing individual video video recording (PVR, Personal Video Recorder) demand of intelligent television product of function progressively strengthens, also be no longer simply to record and playback to the requirement of PVR function, and need more distinctive function.
Based on hard disk (HDD) though PVR function be subject to user's welcome, but along with recorded video quality is higher, limited hard drive space cannot the too many video of typing, cloud (Cloud) service is combined with PVR, there is no the restriction of this respect, and Cloud PVR can make user watch freely recorded program by other-end, thereby can bring better user to experience.
Football is as one of sports the most widely in the world, all can there be every year thousands of all kinds of league matches, Cup, every game all at least continues 90 minutes, if all matches are recorded, what this video library will be very so is huge, and people in most cases just appreciate the wonderful of relevant match simultaneously, therefore only store the wonderful in football match, can, in meeting people's demand, save greatly carrying cost; This wonderful is called video frequency abstract.
But existing video frequency abstract generation technique does not combine with Cloud PVR technology, the Cloud PVR function of intelligent television is relatively simple, does not consider that real-time video frequency abstract generates and is stored in high in the clouds, takes at any time, the application scenarios of real-time sharing.
Summary of the invention
The invention provides a kind of abstract of football video generation method based on cloud TV, real-time video summary can be combined with Cloud PVR technology, alleviate network and high in the clouds pressure.
The present invention also provides a kind of abstract of football video generating apparatus based on cloud TV, real-time video summary can be combined with Cloud PVR technology, alleviates network and high in the clouds pressure.
Technical scheme of the present invention is achieved in that
A kind of abstract of football video generation method based on cloud TV, comprising: football video is carried out to real-time excellent degree analysis, determine excellent video segment, excellent video segment is uploaded to high in the clouds, form video frequency abstract.
An abstract of football video generating apparatus based on cloud TV, comprising:
Excellent video identification module, for football video being carried out to real-time excellent degree analysis, determines excellent video segment;
Transmission module on excellent video, for excellent video segment is uploaded to high in the clouds, forms video frequency abstract.
Visible, the abstract of football video based on cloud TV that the present invention proposes generates method and apparatus, determines in real time the wonderful of football video, and the wonderful of determining is uploaded to high in the clouds, form the video frequency abstract of football video, thereby alleviated network and high in the clouds pressure.
Brief description of the drawings
Fig. 1 is the overall flow schematic diagram that the present invention generates the method example of abstract of football video;
Fig. 2 is the realization flow figure of the embodiment of the present invention one;
Fig. 3 is the abstract of football video generating apparatus structural representation based on cloud TV that the present invention proposes.
Embodiment
The present invention proposes a kind of abstract of football video generation method based on cloud TV, comprising: football video is carried out to real-time excellent degree analysis, determine excellent video segment, excellent video segment is uploaded to high in the clouds, form video frequency abstract.
The present invention has realized the real-time generation of abstract of football video, to the video flowing of recording, taking frame as unit, carry out place detection, mass-tone extraction and Region Segmentation, on the basis of similarity between picture group (GOP, Group of Pictures), carry out forbidden zone line judgement; And this video segment is carried out to audio frequency extraction, and use the support vector machine classifier training to Audio feature analysis, to determine that whether this audio frequency is as excellent audio section, determine by audio frequency and video highlight degree whether this video is excellent video segment.
If Fig. 1 is the overall flow schematic diagram that the present invention generates the method example of abstract of football video.Overall flow comprises: in real time receiver, video stream, the analysis of video highlight degree, the analysis of audio frequency excellence degree, generating video wonderful, upload high in the clouds in real time.
Wherein, spend the analysis phase at video highlight, mainly comprise the following steps:
1) football match place screening: by the image border of Canny operator extraction key frame, determine according to specific rule whether this image is arena map.
2) mass-tone of carrying out football match place is extracted: key frame is carried out to lower-level vision feature extraction, by creating the two-dimensional histogram based on BGR (Blue-Green-Red) color space, get the BGR passage color-values that its proportion is higher; Obtain the court mass-tone information of this match by the BGR passage color value merger of some key frames.
3) region, competition area is cut apart: obtaining on the basis of video mass-tone, the frame of needs analysis is carried out to sub-block to be cut apart, the mass-tone information of each sub-block being carried out two-dimensional histogram analysis and obtained sub-block, then contrasts with video mass-tone information, if information is unanimously thought arena map; Otherwise think additional scene (as auditorium or other camera lenses), by BGR (0,0,0) look, this piece is filled; Former figure after treatment is corroded to expansion denoising, then carry out mask processing with former figure, obtain final place figure.
4) locate the similarity between GOP based on perception hash algorithm: the frame sequence after using perception hash algorithm to place Region Segmentation carries out similarity analysis, if the similarity duration more than GOP, can be carried out forbidden zone line identification on this basis at two.
5) forbidden zone line identification: on the basis of cutting apart in place, use adaptive Canny operator to carry out rim detection, obtain the obvious two-value figure in edge; Utilize all straight lines that occur in Hough linear transformation fitted figure, and determine in figure, whether to contain forbidden zone line by the linear restriction relation between straight line and straight line.
By said process, determine that the video segment that video GOP is similar and comprise forbidden zone line is the video segment that meets video highlight degree; Afterwards, the audio frequency of this video is carried out to excellent degree analysis, concrete mode is: collect in advance the high and low different audio frequency of excellent degree, carry out feature extraction, produce svm classifier device through training, auxiliary video analysis generates excellent video frequency abstract.
Further, because the video of recording is TS stream encapsulation format, for the video frequency abstract that ensures to generate pauses without card, need carry out audio frequency and video PTS, the DTS of each wonderful and PCR, the CC of TS layer and revise.
Based on recording while the strategy of analyzing, to the video flowing of real-time reception, carries out excellent degree analysis according to above-mentioned steps, generating video summary fragment is also called the SDK of cloud platform self, carries out uploading in real time of video frequency abstract.
From said process, the present invention has reduced the complexity that football video is analyzed, and real-time is high; The present invention combines Cloud PVR and video frequency abstract, and the summary generating is in real time transmitted to high in the clouds, reduces network and high in the clouds pressure, is conducive to user and experiences, and has increased product function, has improved product competitiveness.
Below in conjunction with accompanying drawing, lift specific embodiment and introduce in detail.
Embodiment mono-:
The present embodiment introduction generates a concrete example of abstract of football video, as the realization flow figure that Fig. 2 is the present embodiment, for the football video of real-time reception, carries out following steps:
Step 201: the mass-tone that completes competition area is extracted.Comprise following 2 steps:
1) place of carrying out football video is detected
The mass-tone of football video is football pitch ground colour, in mass-tone is extracted, need to filter target frame, removes and does not meet regular frame.Concrete filtering rule is that use Canny algorithm extracts the marginal information of image, and image is divided into N*N piece, adds up the marginal point number in each fritter; By the threshold value T of marginal point number in piece
1be made as empirical value, establish marginal point number in figure and be less than threshold value T
1ratio be R, when R is greater than threshold value T
2time, think and can carry out mass-tone extraction by the place figure that this figure contains certain ratio; Wherein T
2for adaptive threshold.
2) mass-tone extraction is carried out in the competition area that detect
To the arena map detecting, reduce dimension processing, set up two-dimensional histogram, on the basis of this histogram equalization, obtain the mass-tone information of this figure; Mass-tone information by several courts is collected screening, obtains the court mass-tone of this match.
Step 202: start to carry out the similarity analysis of video from current GOP, if similarity is high, continue execution step 203; Otherwise, using next GOP as current GOP, return and carry out this step.
Particularly, the target of Video similarity analysis is the individual non-key frame of N (being generally fps/5) in key frame and this GOP in GOP, and after this non-key frame is analyzed, follow-up B, P frame are not analyzed, and wait for the arrival of next GOP data.
If the key frame of current GOP very similar to non-key frame (being Hamming distance <=3), and very similar to the frame (key frame and N non-key frame) of next GOP, think that similarity is high.
In addition, before carrying out similarity analysis, the frame of needs analysis can be carried out to place Region Segmentation, thereby reduce computation complexity.Can carry out place Region Segmentation according to court mass-tone, concrete mode is:
The frame of needs analysis is carried out to N*N and cut apart, the sub-block after cutting apart is carried out the analysis of BGR two-dimensional histogram equally, thereby obtains the mass-tone information of each sub-block; The mass-tone sequence information of the mass-tone of sub-block and video is contrasted, if meet certain logical condition, think that this sub-block is arena map, otherwise think interfere information (pointing to auditorium or other scenes), use BGR (0,0,0) this sub-block is filled, reduce the interference of subsequent analysis.After Region Segmentation, carry out burn into expansion denoising, carry out mask processing with former figure, be just final place figure.
Step 203: the I frame at the 3rd GOP carries out Hough linear transformation, determine whether forbidden zone line, if had, determine that the fragment of three GOP of first GOP to the meets video highlight degree, and till this fragment that meets video highlight degree will last till forbidden zone heading line off; Otherwise, using next GOP as current GOP, return to execution step 202.
Wherein, carrying out according to Hough linear transformation mode that forbidden zone line judges can be as:
The Performance Area area image of having cut apart is done to smoothing processing, and burn into expands to eliminate noise; The lower-level vision feature of dependency graph picture is set up adaptive Canny boundary operator, obtains obvious bianry image; Then all straight lines that occur in the Hough transformation fitted figure based on probability.Because forbidden zone line may be truncated carrying out when place is cut apart, need carry out merger to the straight line data after matching, and determine whether as forbidden zone line according to the mutual restriction relation between the line of forbidden zone; The restriction relation of forbidden zone line has been in particular in two groups of forbidden zone line parallels or approximate parallel, and slope meets certain logical relation.
Step 204: carry out the analysis of audio frequency excellence degree to meeting the judgement of video highlight degree, judge whether this fragment meets audio frequency excellence degree, if met, determine that this fragment is excellent video; Otherwise.Using next GOP as current GOP, return to execution step 202.
Wherein, the mode of the analysis of audio frequency excellence degree can be:
From ready audio repository, extract in advance the audio frequency characteristics such as Mel cepstrum coefficient (MFCC), color character (Chroma-based feature), zero-crossing rate, utilize support vector machine classifier to carry out the analysis of audio frequency excellence degree.
Step 205: excellent video segment is merged into video frequency abstract, and is uploaded to high in the clouds.
Particularly, video frequency abstract is made up of excellent video segment, the start frame of each video segment is key frame, revises CC, the PCR of TS layer and the PTS of ES layer, DTS information since second video segment, and the SDK that amended excellent video segment is called to cloud platform uploads in real time.
More than introduced the abstract of football video generation method based on cloud TV that the present invention proposes, the present invention also proposes a kind of abstract of football video generating apparatus based on cloud TV, as the structural representation that Fig. 3 is this device, comprising:
Excellent video identification module 310, for football video being carried out to real-time excellent degree analysis, determines excellent video segment;
Transmission module 320 on excellent video, for excellent video segment is uploaded to high in the clouds, forms video frequency abstract.
Above-mentioned excellent video identification module 310 can comprise:
Video highlight degree judge module 311, carries out the place of football video and detects for the distributed intelligence based on edge algorithms and marginal point, determine the key frame that can extract for court mass-tone; The described key frame that can extract for court mass-tone is gathered and carries out court mass-tone extraction according to color histogram and multiframe information; Carry out place Region Segmentation according to court mass-tone information; On the basis of cutting apart in Performance Area territory, according to perception hash algorithm, the GOP similarity of football video is analyzed; On the similar basis of video GOP, carry out forbidden zone line judgement according to the linear feature of the conversion of Hough line and forbidden zone line, determine the video segment that meets video highlight degree;
Audio frequency excellence degree judge module 312, for the video segment for meeting video highlight degree, judges whether corresponding audio frequency meets audio frequency excellence degree; If met, determine that this video segment is excellent video segment.
The distributed intelligence of above-mentioned video highlight degree judge module 311 based on edge algorithms and marginal point carried out the place of football video and detected, and determines that the mode of the key frame that can extract for court mass-tone can be:
The marginal information that uses Canny operator extraction key frame, is divided into N*N piece by key frame, adds up the number of each middle marginal point, and wherein, N is predefined integer;
When the piece that is less than predefined thresholding when contained marginal point accounts for the ratio of all and is greater than predefined ratio, judge that this key frame can extract for court mass-tone.
Above-mentioned video highlight degree judge module 311 gathers the mode of carrying out court mass-tone extraction to the key frame that can extract for court mass-tone according to color histogram and multiframe information:
For the key frame that can extract for court mass-tone, create the two-dimensional histogram based on BGR color space, get the most much higher BGR color value of proportion, and record this color proportion; Many key frames are carried out to mass-tone extraction, merger BGR color value and this color proportion; Get much higher BRG color value of proportion as court mass-tone.
The mode that above-mentioned video highlight degree judge module 311 carries out place Region Segmentation according to court mass-tone information can be:
Frame of video is divided into N*N piece, and wherein, N is predefined integer;
Each piece is carried out to the analysis of BGR two-dimensional histogram, obtain the mass-tone information of each piece; The mass-tone information of each piece is contrasted with the mass-tone sequence information of frame of video respectively; If meet predefined logical condition, judge that this piece is arena map; Otherwise, judge that this piece is interfere information;
Use BGR (0,0,0) to fill the piece that is judged to be interfere information.
Above-mentioned video highlight degree judge module 311 is on the similar basis of video GOP, and carrying out according to the linear feature of Hough line conversion and forbidden zone line mode that forbidden zone line judges can be as:
The Performance Area area image of having cut apart is done to smoothing processing, and burn into expands to eliminate noise; The lower-level vision feature of dependency graph picture is set up adaptive Canny boundary operator, obtains obvious bianry image; The all straight lines that occur in Hough transformation fitted figure based on probability; Straight line data after matching are carried out to merger, and determine whether as forbidden zone line according to the linear restriction relation between the line of forbidden zone.
Above-mentioned audio frequency excellence degree judge module 312 judges that the mode whether corresponding audio frequency meets audio frequency excellence degree is:
Preserve in advance audio sample storehouse, from audio sample storehouse, extract audio frequency characteristics, described audio frequency characteristics comprises Mel cepstrum coefficient MFCC, color character and zero-crossing rate; Utilize support vector machine classifier train and classify, judge whether audio frequency meets audio frequency excellence degree.
As fully visible, the abstract of football video based on cloud TV that the present invention proposes generates method and apparatus, and Cloud PVR and real-time video summarization generation combine, and reduces network and high in the clouds pressure.For the shortcoming that abstract of football video is analyzed frame by frame, computation complexity is high, the present invention only analyzes the key frame in GOP and some non-key frames, in limited information source, excavate fully event information, and extracting on the visual signature of image, reduce computation complexity by modes such as dimensionality reductions.The present invention is directed to the video flowing of recording, taking frame as unit, carry out place detection, mass-tone extraction and Region Segmentation, on the basis of similarity between GOP, carry out forbidden zone line judgement; And this video segment is carried out to audio frequency extraction, and use the support vector machine classifier training to Audio feature analysis, to determine that whether this audio frequency is as excellent audio section, determine by audio frequency and video highlight degree whether this video is excellent video segment.
The foregoing is only preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any amendment of making, be equal to replacement, improvement etc., within all should being included in the scope of protection of the invention.
Claims (14)
1. the abstract of football video generation method based on cloud TV, is characterized in that, the method comprises: football video is carried out to real-time excellent degree analysis, determine excellent video segment, excellent video segment is uploaded to high in the clouds, form video frequency abstract.
2. method according to claim 1, is characterized in that, described football video is carried out to real-time excellent degree analysis, determines that the mode of excellent video segment is:
Distributed intelligence based on edge algorithms and marginal point is carried out the place of football video and is detected, and determines the key frame that can extract for court mass-tone;
The described key frame that can extract for court mass-tone is gathered and carries out court mass-tone extraction according to color histogram and multiframe information;
Carry out place Region Segmentation according to court mass-tone information;
On the basis of cutting apart in Performance Area territory, according to perception hash algorithm, the picture group GOP similarity of football video is analyzed; On the similar basis of video GOP, carry out forbidden zone line judgement according to the linear feature of the conversion of Hough line and forbidden zone line, determine the video segment that meets video highlight degree;
For the video segment that meets video highlight degree, judge whether corresponding audio frequency meets audio frequency excellence degree; If met, determine that this video segment is excellent video segment.
3. method according to claim 2, is characterized in that, the described distributed intelligence based on edge algorithms and marginal point is carried out the place of football video and detected, and determines that the mode of the key frame that can extract for court mass-tone is:
The marginal information that uses Canny operator extraction key frame, is divided into N*N piece by key frame, adds up the number of each middle marginal point, and wherein, N is predefined integer;
When the piece that is less than predefined thresholding when contained marginal point accounts for the ratio of all and is greater than predefined ratio, judge that this key frame can extract for court mass-tone.
4. method according to claim 2, is characterized in that, describedly the key frame that can extract for court mass-tone is gathered to the mode of carrying out court mass-tone extraction according to color histogram and multiframe information is:
For the key frame that can extract for court mass-tone, create the two-dimensional histogram based on BGR color space, get the most much higher BGR color value of proportion, and record this color proportion; Many key frames are carried out to mass-tone extraction, merger BGR color value and this color proportion; Get much higher BRG color value of proportion as court mass-tone.
5. method according to claim 2, is characterized in that, described mode of carrying out place Region Segmentation according to court mass-tone information is:
Frame of video is divided into N*N piece, and wherein, N is predefined integer;
Each piece is carried out to the analysis of BGR two-dimensional histogram, obtain the mass-tone information of each piece; The mass-tone information of each piece is contrasted with the mass-tone sequence information of frame of video respectively; If meet predefined logical condition, judge that this piece is arena map; Otherwise, judge that this piece is interfere information;
Use BGR (0,0,0) to fill the piece that is judged to be interfere information.
6. method according to claim 2, is characterized in that, described on the similar basis of video GOP, according to the linear feature of Hough line conversion and forbidden zone line carry out mode that forbidden zone line judges as:
The Performance Area area image of having cut apart is done to smoothing processing, and burn into expands to eliminate noise; The lower-level vision feature of dependency graph picture is set up adaptive Canny boundary operator, obtains obvious bianry image; The all straight lines that occur in Hough transformation fitted figure based on probability; Straight line data after matching are carried out to merger, and determine whether as forbidden zone line according to the linear restriction relation between the line of forbidden zone.
7. method according to claim 2, is characterized in that, the described mode that judges whether corresponding audio frequency meets audio frequency excellence degree is:
Preserve in advance audio sample storehouse, from audio sample storehouse, extract audio frequency characteristics, described audio frequency characteristics comprises Mel cepstrum coefficient MFCC, color character and zero-crossing rate; Utilize support vector machine classifier train and classify, judge whether audio frequency meets audio frequency excellence degree.
8. the abstract of football video generating apparatus based on cloud TV, is characterized in that, this device comprises:
Excellent video identification module, for football video being carried out to real-time excellent degree analysis, determines excellent video segment;
Transmission module on excellent video, for excellent video segment is uploaded to high in the clouds, forms video frequency abstract.
9. device according to claim 8, is characterized in that, described excellent video identification module comprises:
Video highlight degree judge module, carries out the place of football video and detects for the distributed intelligence based on edge algorithms and marginal point, determine the key frame that can extract for court mass-tone; The described key frame that can extract for court mass-tone is gathered and carries out court mass-tone extraction according to color histogram and multiframe information; Carry out place Region Segmentation according to court mass-tone information; On the basis of cutting apart in Performance Area territory, according to perception hash algorithm, the GOP similarity of football video is analyzed; On the similar basis of video GOP, carry out forbidden zone line judgement according to the linear feature of the conversion of Hough line and forbidden zone line, determine the video segment that meets video highlight degree;
Audio frequency excellence degree judge module, for the video segment for meeting video highlight degree, judges whether corresponding audio frequency meets audio frequency excellence degree; If met, determine that this video segment is excellent video segment.
10. device according to claim 8, is characterized in that, the distributed intelligence of described video highlight degree judge module based on edge algorithms and marginal point carried out the place of football video and detected, and determines that the mode of the key frame that can extract for court mass-tone is:
The marginal information that uses Canny operator extraction key frame, is divided into N*N piece by key frame, adds up the number of each middle marginal point, and wherein, N is predefined integer;
When the piece that is less than predefined thresholding when contained marginal point accounts for the ratio of all and is greater than predefined ratio, judge that this key frame can extract for court mass-tone.
11. devices according to claim 8, is characterized in that, described video highlight degree judge module gathers the mode of carrying out court mass-tone extraction to the key frame that can extract for court mass-tone according to color histogram and multiframe information and is:
For the key frame that can extract for court mass-tone, create the two-dimensional histogram based on BGR color space, get the most much higher BGR color value of proportion, and record this color proportion; Many key frames are carried out to mass-tone extraction, merger BGR color value and this color proportion; Get much higher BRG color value of proportion as court mass-tone.
12. devices according to claim 8, is characterized in that, the mode that described video highlight degree judge module carries out place Region Segmentation according to court mass-tone information is:
Frame of video is divided into N*N piece, and wherein, N is predefined integer;
Each piece is carried out to the analysis of BGR two-dimensional histogram, obtain the mass-tone information of each piece; The mass-tone information of each piece is contrasted with the mass-tone sequence information of frame of video respectively; If meet predefined logical condition, judge that this piece is arena map; Otherwise, judge that this piece is interfere information;
Use BGR (0,0,0) to fill the piece that is judged to be interfere information.
13. devices according to claim 8, is characterized in that, described video highlight degree judge module on the similar basis of video GOP, according to the linear feature of Hough line conversion and forbidden zone line carry out mode that forbidden zone line judges as:
The Performance Area area image of having cut apart is done to smoothing processing, and burn into expands to eliminate noise; The lower-level vision feature of dependency graph picture is set up adaptive Canny boundary operator, obtains obvious bianry image; The all straight lines that occur in Hough transformation fitted figure based on probability; Straight line data after matching are carried out to merger, and determine whether as forbidden zone line according to the linear restriction relation between the line of forbidden zone.
14. devices according to claim 8, is characterized in that, described audio frequency excellence degree judge module judges that the mode whether corresponding audio frequency meets audio frequency excellence degree is:
Preserve in advance audio sample storehouse, from audio sample storehouse, extract audio frequency characteristics, described audio frequency characteristics comprises Mel cepstrum coefficient MFCC, color character and zero-crossing rate; Utilize support vector machine classifier train and classify, judge whether audio frequency meets audio frequency excellence degree.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410401744.4A CN104123396B (en) | 2014-08-15 | 2014-08-15 | A kind of abstract of football video generation method and device based on cloud TV |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410401744.4A CN104123396B (en) | 2014-08-15 | 2014-08-15 | A kind of abstract of football video generation method and device based on cloud TV |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104123396A true CN104123396A (en) | 2014-10-29 |
CN104123396B CN104123396B (en) | 2017-07-07 |
Family
ID=51768807
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410401744.4A Active CN104123396B (en) | 2014-08-15 | 2014-08-15 | A kind of abstract of football video generation method and device based on cloud TV |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104123396B (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106937160A (en) * | 2017-03-28 | 2017-07-07 | 青岛海信电器股份有限公司 | Display methods and device of a kind of recorded file for terminal in fast forwarding and fast rewinding |
CN106993209A (en) * | 2016-01-20 | 2017-07-28 | 上海慧体网络科技有限公司 | A kind of method that short video clip is carried out based on mobile terminal technology |
CN108093311A (en) * | 2017-12-28 | 2018-05-29 | 广东欧珀移动通信有限公司 | Processing method, device, storage medium and the electronic equipment of multimedia file |
CN108615043A (en) * | 2016-12-12 | 2018-10-02 | 中移(杭州)信息技术有限公司 | A kind of video classification methods and system |
CN108683924A (en) * | 2018-05-30 | 2018-10-19 | 北京奇艺世纪科技有限公司 | A kind of method and apparatus of video processing |
CN108882003A (en) * | 2018-07-25 | 2018-11-23 | 安徽新华学院 | A kind of electronic software control system that can detect excellent race automatically |
CN109344697A (en) * | 2018-08-16 | 2019-02-15 | 中国科学院信息工程研究所 | A kind of Wonderful time recognition methods in antagonism contest |
CN109691124A (en) * | 2016-06-20 | 2019-04-26 | 皮克索洛特公司 | For automatically generating the method and system of Video Highlights |
CN110381367A (en) * | 2019-07-10 | 2019-10-25 | 咪咕文化科技有限公司 | A kind of method for processing video frequency, equipment and computer readable storage medium |
CN110505521A (en) * | 2019-08-28 | 2019-11-26 | 咪咕动漫有限公司 | A kind of live streaming match interactive approach, electronic equipment, storage medium and system |
CN113052033A (en) * | 2021-03-15 | 2021-06-29 | 深圳软牛科技有限公司 | Video wonderness measuring method, device, equipment and storage medium |
CN113052149A (en) * | 2021-05-20 | 2021-06-29 | 平安科技(深圳)有限公司 | Video abstract generation method and device, computer equipment and medium |
US11627383B2 (en) | 2018-04-10 | 2023-04-11 | Samsung Electronics Co., Ltd. | Electronic device and operation method thereof |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070109446A1 (en) * | 2005-11-15 | 2007-05-17 | Samsung Electronics Co., Ltd. | Method, medium, and system generating video abstract information |
CN101013444A (en) * | 2007-02-13 | 2007-08-08 | 华为技术有限公司 | Method and apparatus for adaptively generating abstract of football video |
CN101431689A (en) * | 2007-11-05 | 2009-05-13 | 华为技术有限公司 | Method and device for generating video abstract |
CN101778257A (en) * | 2010-03-05 | 2010-07-14 | 北京邮电大学 | Generation method of video abstract fragments for digital video on demand |
US20100284670A1 (en) * | 2008-06-30 | 2010-11-11 | Tencent Technology (Shenzhen) Company Ltd. | Method, system, and apparatus for extracting video abstract |
CN102801978A (en) * | 2012-08-08 | 2012-11-28 | 浪潮集团有限公司 | System for video data acquisition and management based on cloud storage |
-
2014
- 2014-08-15 CN CN201410401744.4A patent/CN104123396B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070109446A1 (en) * | 2005-11-15 | 2007-05-17 | Samsung Electronics Co., Ltd. | Method, medium, and system generating video abstract information |
CN101013444A (en) * | 2007-02-13 | 2007-08-08 | 华为技术有限公司 | Method and apparatus for adaptively generating abstract of football video |
CN101431689A (en) * | 2007-11-05 | 2009-05-13 | 华为技术有限公司 | Method and device for generating video abstract |
US20100284670A1 (en) * | 2008-06-30 | 2010-11-11 | Tencent Technology (Shenzhen) Company Ltd. | Method, system, and apparatus for extracting video abstract |
CN101778257A (en) * | 2010-03-05 | 2010-07-14 | 北京邮电大学 | Generation method of video abstract fragments for digital video on demand |
CN102801978A (en) * | 2012-08-08 | 2012-11-28 | 浪潮集团有限公司 | System for video data acquisition and management based on cloud storage |
Cited By (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106993209A (en) * | 2016-01-20 | 2017-07-28 | 上海慧体网络科技有限公司 | A kind of method that short video clip is carried out based on mobile terminal technology |
CN109691124A (en) * | 2016-06-20 | 2019-04-26 | 皮克索洛特公司 | For automatically generating the method and system of Video Highlights |
CN108615043A (en) * | 2016-12-12 | 2018-10-02 | 中移(杭州)信息技术有限公司 | A kind of video classification methods and system |
CN108615043B (en) * | 2016-12-12 | 2021-05-25 | 中移(杭州)信息技术有限公司 | Video classification method and system |
CN106937160A (en) * | 2017-03-28 | 2017-07-07 | 青岛海信电器股份有限公司 | Display methods and device of a kind of recorded file for terminal in fast forwarding and fast rewinding |
CN108093311A (en) * | 2017-12-28 | 2018-05-29 | 广东欧珀移动通信有限公司 | Processing method, device, storage medium and the electronic equipment of multimedia file |
CN108093311B (en) * | 2017-12-28 | 2021-02-02 | Oppo广东移动通信有限公司 | Multimedia file processing method and device, storage medium and electronic equipment |
US11627383B2 (en) | 2018-04-10 | 2023-04-11 | Samsung Electronics Co., Ltd. | Electronic device and operation method thereof |
CN108683924A (en) * | 2018-05-30 | 2018-10-19 | 北京奇艺世纪科技有限公司 | A kind of method and apparatus of video processing |
CN108882003A (en) * | 2018-07-25 | 2018-11-23 | 安徽新华学院 | A kind of electronic software control system that can detect excellent race automatically |
CN109344697B (en) * | 2018-08-16 | 2021-11-09 | 中国科学院信息工程研究所 | Method for identifying wonderful moment in antagonism competition |
CN109344697A (en) * | 2018-08-16 | 2019-02-15 | 中国科学院信息工程研究所 | A kind of Wonderful time recognition methods in antagonism contest |
CN110381367A (en) * | 2019-07-10 | 2019-10-25 | 咪咕文化科技有限公司 | A kind of method for processing video frequency, equipment and computer readable storage medium |
CN110505521B (en) * | 2019-08-28 | 2021-11-23 | 咪咕动漫有限公司 | Live broadcast competition interaction method, electronic equipment, storage medium and system |
CN110505521A (en) * | 2019-08-28 | 2019-11-26 | 咪咕动漫有限公司 | A kind of live streaming match interactive approach, electronic equipment, storage medium and system |
CN113052033A (en) * | 2021-03-15 | 2021-06-29 | 深圳软牛科技有限公司 | Video wonderness measuring method, device, equipment and storage medium |
CN113052149A (en) * | 2021-05-20 | 2021-06-29 | 平安科技(深圳)有限公司 | Video abstract generation method and device, computer equipment and medium |
CN113052149B (en) * | 2021-05-20 | 2021-08-13 | 平安科技(深圳)有限公司 | Video abstract generation method and device, computer equipment and medium |
Also Published As
Publication number | Publication date |
---|---|
CN104123396B (en) | 2017-07-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104123396A (en) | Soccer video abstract generation method and device based on cloud television | |
Perez et al. | Detection of real-world fights in surveillance videos | |
US20210042960A1 (en) | Computer vision on broadcast video | |
KR102433393B1 (en) | Apparatus and method for recognizing character in video contents | |
KR101967086B1 (en) | Entity-based temporal segmentation of video streams | |
CA2761187C (en) | Systems and methods for the autonomous production of videos from multi-sensored data | |
CN106162223B (en) | News video segmentation method and device | |
US8750681B2 (en) | Electronic apparatus, content recommendation method, and program therefor | |
KR100873638B1 (en) | Image processing method and apparatus | |
CN108600865B (en) | A kind of video abstraction generating method based on super-pixel segmentation | |
US20160148650A1 (en) | Video processing system with custom chaptering and methods for use therewith | |
CN103200463A (en) | Method and device for generating video summary | |
US20150169960A1 (en) | Video processing system with color-based recognition and methods for use therewith | |
JP2008312215A (en) | Video-image analyzer, video-image analyzing method, automatic digest preparation system, and automatic highlight extraction system | |
US10181083B2 (en) | Scene change detection and logging | |
Wang et al. | Affection arousal based highlight extraction for soccer video | |
CN105657310B (en) | A kind of mobile phone screen enrolls method in real time | |
Ramos et al. | Fast-forward video based on semantic extraction | |
CN103187083B (en) | A kind of storage means based on time domain video fusion and system thereof | |
Husa et al. | HOST-ATS: automatic thumbnail selection with dashboard-controlled ML pipeline and dynamic user survey | |
CN111738042A (en) | Identification method, device and storage medium | |
KR102426089B1 (en) | Electronic device and Method for generating summary image of electronic device | |
Salehin et al. | Human visual field based saliency prediction method using Eye Tracker data for video summarization | |
CN105989063B (en) | Video retrieval method and device | |
CN106162092B (en) | A kind of method and system sport video acquisition and played |
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 |