CN109635707A - A kind of video lens extracting method based on feature identification - Google Patents
A kind of video lens extracting method based on feature identification Download PDFInfo
- Publication number
- CN109635707A CN109635707A CN201811486550.3A CN201811486550A CN109635707A CN 109635707 A CN109635707 A CN 109635707A CN 201811486550 A CN201811486550 A CN 201811486550A CN 109635707 A CN109635707 A CN 109635707A
- Authority
- CN
- China
- Prior art keywords
- time point
- video
- feature
- present frame
- feature identification
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Studio Devices (AREA)
Abstract
The invention discloses a kind of video lens extracting methods based on feature identification, comprising: obtains current time, and judges whether current time is time point in preset time point set, if so, obtaining the present frame of target video；Feature identification is carried out to the involved party in the corresponding picture of present frame according to characteristic recognition method, obtains the feature of involved party；Judge whether the feature of involved party matches with preset target signature, if so, intercepting the continuous video frame of preset quantity before and after present frame as target video camera lens；Judge whether current time is the last one time point in preset time point set, if so, output target video camera lens.
Description
Technical field
The present invention relates to technical field of video processing more particularly to a kind of video lens extraction sides based on feature identification
Method.
Background technique
Broadcasting and TV media possess resource abundant, for example, having a large amount of professional sports race signal can be the Internet media
Content support is provided, and competitive sports are also the topic usually continued saying it with interest in people's life, possess a large amount of audient；Internet
Media have the characteristics that spread speed is fast, disseminate wide, information fragmentation, can continue to keep temperature, on the one hand, can push
On the other hand the production of broadcasting and TV media also proposes the requirement of content and efficiency to the existing manufacturing system of broadcasting and TV media.
When existing information is issued, bout generally has one or two hour, and the time span of content is unfavorable for interconnecting
Net, social media are propagated, the development of internet and amalgamation media, so that more and more people adapt to and the letter of demand " fragmentation "
Breath.Therefore, the mode that match Highlight is directed to internet publication again of extracting generally is taken, secondly, existing extraction match essence
The method of color camera lens is substantially got ready manually after video includes completion, is manually sheared, it is clear that such method is not
A large amount of manpower and time are only consumed, and efficiency is very low, can not meet the needs of timeliness.
Summary of the invention
Technical problems based on background technology, the invention proposes a kind of video lens extractions based on feature identification
Method；
A kind of video lens extracting method based on feature identification proposed by the present invention, comprising:
S1, current time is obtained, and judges whether current time is time point in preset time point set, if so,
Obtain the present frame of target video；
S2, feature identification is carried out to the involved party in the corresponding picture of present frame according to characteristic recognition method, obtains behavior
The feature of people；
S3, judge whether the feature of involved party matches with preset target signature, if so, intercepting before and after present frame default
The continuous video frame of quantity executes step S4 as target video camera lens；Otherwise, step S4 is executed；
S4, judge whether current time is the last one time point in preset time point set, if so, output target
Video lens；If it is not, executing step S1.
Preferably, step S2 is specifically included:
Face characteristic identification is carried out to the involved party in the corresponding picture of present frame according to face characteristic recognition methods, is obtained
The face characteristic of involved party.
Preferably, in step S3, the target signature is specifically included: indicating the people that happy face characteristic, instruction are shed tears
The face characteristic that face feature, instruction are hailed.
Preferably, step S2 is specifically included:
Motion characteristic identification is carried out to the involved party in the corresponding picture of present frame according to motion characteristic recognition methods, is obtained
The motion characteristic of involved party.
Preferably, in step S3, the target signature is specifically included: indicating that the motion characteristic hailed, instruction are celebrated dynamic
Make feature.
Preferably, the time point ascending order arrangement in step S1, in the preset time point set.
In the present invention, the present frame of target video is obtained, according to characteristic recognition method in the corresponding picture of present frame
Involved party carries out feature identification, obtains the feature of involved party, judges whether the feature of involved party matches with preset target signature,
If so, the continuous video frame for intercepting preset quantity before and after present frame identifies video in this way, passing through as target video camera lens
Face characteristic and motion characteristic in frame picture, the automatic featured videos camera lens searched when scoring, winning, operation is simple and reliable, effect
Rate is high, does not need professional video editing technical ability, can effectively extract the excellent picture in sports tournament video.
Detailed description of the invention
Fig. 1 is a kind of flow diagram of video lens extracting method based on feature identification proposed by the present invention.
Specific embodiment
Referring to Fig.1, a kind of video lens extracting method based on feature identification proposed by the present invention, comprising:
Step S1 obtains current time, and judges whether current time is time point in preset time point set, if
It is the present frame for obtaining target video.
Time point ascending order arrangement in this step, in the preset time point set.
In concrete scheme, continual acquisition current time, if current time is the time point in time point set
When, obtain target video when former frame picture, if current time is not any one time point in time point set, after
Continue continual acquisition current time.
Further, the time point ascending order arrangement in time point set, by the last one time point in time point set
It is set as end time point, when current time reaches the last one time point in time point set, stops obtaining target view
The present frame of frequency.
Step S2 carries out feature identification to the involved party in the corresponding picture of present frame according to characteristic recognition method, obtains
The feature of involved party.
This step specifically includes: carrying out people to the involved party in the corresponding picture of present frame according to face characteristic recognition methods
The identification of face feature, obtains the face characteristic of involved party.
This step specifically includes: being moved according to motion characteristic recognition methods to the involved party in the corresponding picture of present frame
Make feature identification, obtains the motion characteristic of involved party.
In concrete scheme, face characteristic is carried out to the involved party in the corresponding picture of present frame by face recognition technology
Identification, obtains the face characteristic of involved party；
And/or motion characteristic knowledge is carried out to the involved party in the corresponding picture of present frame by motion characteristic identification technology
Not, the motion characteristic of involved party is obtained.
Step S3, judges whether the feature of involved party matches with preset target signature, if so, intercepting before and after present frame
The continuous video frame of preset quantity executes step S4 as target video camera lens；Otherwise, step S4 is executed.
In this step, target signature includes the happy face characteristic of instruction, indicates that the face characteristic shed tears, instruction are hailed
Face characteristic.
In this step, target signature includes the motion characteristic of the motion characteristic of instruction cheer, instruction celebration.
In concrete scheme, when goal, triumph, celebrating, often the expression of involved party, movement are mostly mutually similar, when working as
The face characteristic of involved party in the corresponding picture of the previous frame face characteristic happy with instruction indicates the face characteristic shed tears, refers to
When showing that one matches in the face characteristic of cheer, illustrate that target video has goal, triumph, the picture celebrated, and/or, when current
One in the motion characteristic of motion characteristic, instruction celebration that the motion characteristic of involved party in the corresponding picture of frame and instruction are hailed
When matching, illustrate that target video has goal, triumph, the picture celebrated；At this point, intercepting the continuous of preset quantity before and after present frame
Video frame as target video camera lens.
Step S4 judges whether current time is the last one time point in preset time point set, if so, output
Target video camera lens；If it is not, executing step S1.
In concrete scheme, when current time is the last one time point in preset time point set, illustrate mesh
Mark videoendoscope head retrieval finishes, and exports all target video camera lenses, that is, exports all Highlights.
In present embodiment, the present frame of target video is obtained, according to characteristic recognition method to the corresponding picture of present frame
In involved party carry out feature identification, obtain the feature of involved party, judge involved party feature whether with preset target signature
Matching, if so, the continuous video frame of interception preset quantity is as target video camera lens before and after present frame, in this way, passing through knowledge
Face characteristic and motion characteristic in other video frame picture, the automatic featured videos camera lens searched when scoring, winning, it is easy to operate can
It leans on, it is high-efficient, professional video editing technical ability is not needed, the excellent picture in sports tournament video can be effectively extracted.
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto,
Anyone skilled in the art in the technical scope disclosed by the present invention, according to the technique and scheme of the present invention and its
Inventive concept is subject to equivalent substitution or change, should be covered by the protection scope of the present invention.
Claims (6)
1. a kind of video lens extracting method based on feature identification characterized by comprising
S1, current time is obtained, and judges whether current time is time point in preset time point set, if so, obtaining
The present frame of target video；
S2, feature identification is carried out to the involved party in the corresponding picture of present frame according to characteristic recognition method, obtains involved party's
Feature；
S3, judge whether the feature of involved party matches with preset target signature, if so, intercepting preset quantity before and after present frame
Continuous video frame as target video camera lens, and execute step S4；Otherwise, step S4 is executed；
S4, judge whether current time is the last one time point in preset time point set, if so, output target video
Camera lens；If it is not, executing step S1.
2. the video lens extracting method according to claim 1 based on feature identification, which is characterized in that step S2, tool
Body includes:
Face characteristic identification is carried out to the involved party in the corresponding picture of present frame according to face characteristic recognition methods, obtains behavior
The face characteristic of people.
3. the video lens extracting method according to claim 2 based on feature identification, which is characterized in that in step S3,
The target signature, specifically includes: face characteristic, the face of instruction cheer for indicating that happy face characteristic, instruction are shed tears are special
Sign.
4. the video lens extracting method according to claim 1 based on feature identification, which is characterized in that step S2, tool
Body includes:
Motion characteristic identification is carried out to the involved party in the corresponding picture of present frame according to motion characteristic recognition methods, obtains behavior
The motion characteristic of people.
5. the video lens extracting method according to claim 4 based on feature identification, which is characterized in that in step S3,
The target signature, specifically includes: indicating the motion characteristic that the motion characteristic hailed, instruction are celebrated.
6. the video lens extracting method according to claim 1 based on feature identification, which is characterized in that in step S1,
Time point ascending order arrangement in the preset time point set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811486550.3A CN109635707A (en) | 2018-12-06 | 2018-12-06 | A kind of video lens extracting method based on feature identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811486550.3A CN109635707A (en) | 2018-12-06 | 2018-12-06 | A kind of video lens extracting method based on feature identification |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109635707A true CN109635707A (en) | 2019-04-16 |
Family
ID=66071554
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811486550.3A Pending CN109635707A (en) | 2018-12-06 | 2018-12-06 | A kind of video lens extracting method based on feature identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109635707A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111832860A (en) * | 2019-04-19 | 2020-10-27 | 杭州海康威视数字技术股份有限公司 | Video-based scoring method and device and electronic equipment |
| CN112632329A (en) * | 2020-12-18 | 2021-04-09 | 咪咕互动娱乐有限公司 | Video extraction method and device, electronic equipment and storage medium |
| CN115761883A (en) * | 2022-11-10 | 2023-03-07 | 南京审计大学 | Automatic field control method of intelligent badminton stadium based on AI technology |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104520875A (en) * | 2012-07-11 | 2015-04-15 | 意大利广播电视股份公司 | A method and an apparatus for the extraction of descriptors from video content, preferably for search and retrieval purpose |
| CN107241645A (en) * | 2017-06-09 | 2017-10-10 | 成都索贝数码科技股份有限公司 | A kind of method that splendid moment of scoring is automatically extracted by the subtitle recognition to video |
| CN107330407A (en) * | 2017-06-30 | 2017-11-07 | 北京金山安全软件有限公司 | Facial expression recognition method and device, electronic equipment and storage medium |
| CN108491764A (en) * | 2018-03-05 | 2018-09-04 | 竹间智能科技（上海）有限公司 | A kind of video human face Emotion identification method, medium and equipment |
-
2018
- 2018-12-06 CN CN201811486550.3A patent/CN109635707A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104520875A (en) * | 2012-07-11 | 2015-04-15 | 意大利广播电视股份公司 | A method and an apparatus for the extraction of descriptors from video content, preferably for search and retrieval purpose |
| CN107241645A (en) * | 2017-06-09 | 2017-10-10 | 成都索贝数码科技股份有限公司 | A kind of method that splendid moment of scoring is automatically extracted by the subtitle recognition to video |
| CN107330407A (en) * | 2017-06-30 | 2017-11-07 | 北京金山安全软件有限公司 | Facial expression recognition method and device, electronic equipment and storage medium |
| CN108491764A (en) * | 2018-03-05 | 2018-09-04 | 竹间智能科技（上海）有限公司 | A kind of video human face Emotion identification method, medium and equipment |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111832860A (en) * | 2019-04-19 | 2020-10-27 | 杭州海康威视数字技术股份有限公司 | Video-based scoring method and device and electronic equipment |
| CN111832860B (en) * | 2019-04-19 | 2023-10-17 | 杭州海康威视数字技术股份有限公司 | Video-based scoring method and device and electronic equipment |
| CN112632329A (en) * | 2020-12-18 | 2021-04-09 | 咪咕互动娱乐有限公司 | Video extraction method and device, electronic equipment and storage medium |
| CN115761883A (en) * | 2022-11-10 | 2023-03-07 | 南京审计大学 | Automatic field control method of intelligent badminton stadium based on AI technology |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11657079B2 (en) | System and method for identifying social trends | |
| CN109635707A (en) | A kind of video lens extracting method based on feature identification | |
| CN106162223B (en) | News video segmentation method and device | |
| US10755086B2 (en) | Picture ranking method, and terminal | |
| US10043551B2 (en) | Techniques to save or delete a video clip | |
| US8503523B2 (en) | Forming a representation of a video item and use thereof | |
| CN109194978A (en) | Live video clipping method, device and electronic equipment | |
| Xu et al. | Pllava: Parameter-free llava extension from images to videos for video dense captioning | |
| CN104199933A (en) | Multi-modal information fusion football video event detection and semantic annotation method | |
| CN104899261B (en) | A kind of apparatus and method for building structuring video image information | |
| CN101720006B (en) | Positioning method suitable for representative frame extracted by video keyframe | |
| CN110188241B (en) | Intelligent manufacturing system and manufacturing method for events | |
| JP2013525916A5 (en) | ||
| CN104298748A (en) | Device and method for face search in videos | |
| WO2016101766A1 (en) | Method and device for obtaining similar face images and face image information | |
| CN103973550A (en) | Method, system and device for rapidly and intelligently identifying instant messaging application ID (identity) number and carrying out instant messaging | |
| CN104063701A (en) | Rapid television station caption recognition system based on SURF vocabulary tree and template matching and implementation method of rapid television station caption recognition system | |
| CN101324919A (en) | Photograph video contrast method | |
| CN103986981B (en) | The recognition methods of the plot fragment of multimedia file and device | |
| CN111770359A (en) | Event video clipping method, system and computer readable storage medium | |
| CN104036227A (en) | Electronic music score generating method and mobile terminal | |
| US8131773B2 (en) | Search information managing for moving image contents | |
| US8896708B2 (en) | Systems and methods for determining, storing, and using metadata for video media content | |
| CN105472407A (en) | Automatic video index and alignment method based on continuous image features | |
| JP7525118B2 (en) | Participant identification in images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |