CN102567720A - Face identification method and face identification device for Flash online video - Google Patents
Face identification method and face identification device for Flash online video Download PDFInfo
- Publication number
- CN102567720A CN102567720A CN2011104447218A CN201110444721A CN102567720A CN 102567720 A CN102567720 A CN 102567720A CN 2011104447218 A CN2011104447218 A CN 2011104447218A CN 201110444721 A CN201110444721 A CN 201110444721A CN 102567720 A CN102567720 A CN 102567720A
- Authority
- CN
- China
- Prior art keywords
- face
- video data
- online video
- face identification
- 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
Images
Landscapes
- Image Processing (AREA)
- Collating Specific Patterns (AREA)
Abstract
The invention provides a face identification method for a Flash online video. The method comprises the steps of calculating the current actual resource use rate of a client; regulating identification parameters according to the current actual resource use rate and sampling the data of an online video according to the regulated identification parameters; and adopting a face identification module to carry out face identification to the currently sampled data of the online video. The invention also provides a face identification device for a Flash online video. The device comprises a processing module, a system interface and a face identification module, wherein the processing module is used for calculating the current actual resource use rate of the client and regulating the identification parameters according to the current actual resource use rate; the system interface is used for sampling the data of the online video according to the regulated identification parameters; and the face identification module is used for carrying out face identification to the currently sampled data of the online video. The face identification method and the face identification device can realize face identification on the data of the online video.
Description
Technical field
The present invention relates to face recognition technology, relate in particular to the face identification method and the face identification device of Flash Online Video.
Background technology
Along with the development of internet Online Video, increasing Internet firm has adopted Flash Online Video form.For example external Youtube, domestic 56, excellent cruel, potato etc. have all adopted this video format, and at present, the overwhelming majority's who sees on the internet video all is based on this video format.On this basis, also there is diversified video interactive mode to occur, such as online live, online video recording, video monitoring etc.
Face recognition module is in the computer graphics, a kind of technology that the human face region in image or the video is discerned, judged.This face recognition module mainly rely on prior art in still constantly by perfect recognition of face information bank, English name: OpenCV (Open Source Computer Vision Library, computer picture increase income framework).Various programming languages utilize this cover identifying information storehouse to realize the function of face recognition module, accomplish the identifying to people's face.
Because very big to the data computation amount of recognition of face, as the Online Video data not to be carried out computing on market face recognition application.And the recognition of face that exists at present, majority is the identification to simple view data.For example the interest figure product of 56.com is exactly a typical application example, and the operation of this product is: the user at first uploads the picture of a band people face; The background process program is after receiving this picture, and OpenCV people's face video library of said corresponding computer development language carries out human face region to image and discerns above the utilization, and the zone that thereupon will recognize passes to image processing program; Image processing program cuts picture, and people's face image data is returned to the front end page, and just user's webpage end that can browse to is realized obtaining people's face.
There is following shortcoming in face recognition technology to video at present: 1. treatment scheme is loaded down with trivial details.The user whenever uploads a pictures, all need pass through and upload, and compression, the recognition of face on the server are judged, pass picture back, show these steps.
2. operational efficiency is low, and range of application is narrow.Because efficiency can't solve, such recognition application can only be used in picture and handle, simultaneously, with the user be a non real-time process alternately, the also just static picture that user experience arrives generates impression.
Summary of the invention
The present invention provides a kind of face identification method and face identification device of Flash Online Video, can realize the recognition of face to the Online Video data.
The face identification method of Flash Online Video provided by the invention, the at first current real resource utilization rate of computing client end; Regulate identification parameter according to said current real resource utilization rate, said Online Video data are sampled according to the identification parameter after regulating; Through face recognition module current Online Video data of sampling are carried out recognition of face.
The face identification device of Flash Online Video provided by the invention comprises processing module, is used for the current real resource utilization rate of computing client end; Regulate identification parameter according to said current real resource utilization rate; System interface is used for according to the identification parameter after regulating said Online Video data being sampled; Face recognition module is carried out recognition of face to current Online Video data of sampling.
The present invention has been applied to face recognition technology in the Flash Online Video field; Having solved in the past can only be to the problem of static images match people face; Let the user can be in seeing any interbehaviors relevant such as video, live or video recording with video match people face in real time, enriched user experience.Improved macrooperation amount the taking of recognition of face to the full extent to low configure user system resource; Through calculating the resources occupation rate of subscription client; Dynamically adjust identification parameter; Thereby the precision of adjustment people face coupling reduces the extra pressure that the recognition of face data computation brings for video playback, makes the user smoothness experience this technical application.The invention enables the Online Video website to develop and more valued added applications experience, for Internet development has been created bigger feeling of freshness and value.
Description of drawings
Fig. 1 is the process flow diagram of embodiment 1;
Fig. 2 is the process flow diagram of step 2 and step 3 among the embodiment 4;
Fig. 3 obtains to preserve the process flow diagram of online video data to the authority of internal memory among the embodiment 5;
Fig. 4 is the theory diagram of face identification device among the embodiment 7.
Embodiment
Embodiment 1: because the present invention hopes to realize discerning the recognition of face in the dynamic video; Therefore need carry out Real-time and Dynamic calculating timely to picture; Suppose that present target video size is 640x480, frame frequency was 30 frame/seconds, just meaned in one second, to carry out 30 times identification process.Can consume a large amount of cpu so undoubtedly calculates.In order to address this problem, the present invention has realized that on base of recognition the client resource current according to the user takies situation, the perhaps demand of some application-specific, and the dynamic adjustments identification parameter guarantees that this function can not have influence on user's video playback efficient.The present invention realizes through following steps:
The Flash playing program has a process and obtains the active client real time data, comprises cpu utilization rate, actual motion frame frequency etc., the actual source utilization rate (step 1) that the present invention is at first current according to these real time data computing client ends; Regulate identification parameter (step 2) according to said current real resource utilization rate, according to the identification parameter after regulating to the said Online Video data (step 3) of sampling; Through face recognition module current Online Video data of sampling are carried out recognition of face (step 4).
As an embodiment; The face recognition module of using in the step 4 can be utilized existing framework of increasing income in the prior art; For example, disclosed among the http://www.libspark.org/browser/as3/Marilena the data migration of the OpenCV recognition of face information bank method to the Flash Actionscript development language.As an embodiment, behind the single frames data capture of step 3, can deposit in the internal memory temporarily and can certainly be stored in other positions video.Just can handle the image data of single frames this moment through the face recognition module that interface is delivered to Flash.As an embodiment, after the real-time operation processing through face recognition module, can concrete data such as the people's face position that is identified be returned with the mechanism of incident bubbling.If in the view data of importing into, do not recognize people's face, can distribute events corresponding yet.
Embodiment 2: in the present embodiment, the identification parameter in the step 2 is the image sampling size of Online Video data.According to current real resource utilization rate, dynamic adjustments is imported the image array of face recognition module in the step 2, the size of the view data that adjustment is sampled.In step 3, the picture size of importing into is big more, and the precision of coupling is high more, but the CPU that consumes is just high more, and it is just long more mate the used time.In the configuration of client comparatively during low side, can realize reducing the normal play that precision guarantees video and play up.
Embodiment 3: in the present embodiment, the identification parameter in the step 2 is the image sampling frequency of Online Video data.According to current real resource utilization rate, dynamic adjustments is imported the SF of face recognition module in the step 2.In step 3, the frequency of sampling is high more, and the smoothness of coupling is just high more, but the CPU that consumes is just big more.In the configuration of client comparatively during low side, can realize reducing the coupling smoothness and guarantee that the normal play of video plays up.
Embodiment 4: as the further improvement to the foregoing description 2 and embodiment 3; The identification parameter of step 2 comprises image sampling size and image sampling frequency simultaneously in the present embodiment; In the step 2 these two kinds of identification parameters are all carried out dynamic adjustments; Fig. 2 is the process flow diagram of present embodiment; When judging that according to the real resource occupancy that calculates being in too much CPU at present consumes, reduce the picture size and the SF of image sampling, according to the image sampling size after reducing with SF is sampled and recognition of face.Otherwise, with the go forward side by side pedestrian's face identification of normal parameter sampled images.
Embodiment 5: in the present embodiment, step 3 just needs to obtain through the system interface of Flash Actionscript the video data of each frame of Flash Online Video for the Online Video data are sampled.Flash API provides a kind of method can the screen-picture data in the screen be saved in the internal memory, and code is following:
BitmapData.draw (screen content of preservation);
In this process, there is a problem; Because video is not the resource at subscriber's local usually; Flash limits aspect security to some extent; Need to preserve the screen-picture data smoothly in internal memory, therefore before execution in step 3, just need at first do the operation of some authorities, make the video that is loaded to be hunted down.Usually have two types of different Flash Online Video agreements:
1.1http type video flow data
Usually our video on video website, seen all is delivered to client through this agreement.For this video, need to load the cross-domain strategy file under the pairing server of video file, video file content is to the authority of internal memory under this server of acquisition preservation.Specific code is following:
Security.loadPolicyFile(″http://127.0.0.1/crossdomain.xml″);
1.2rmpt type video flow data
Such video protocols is corresponding has a media server, such as the Flash Media Server of Adobe company; Be called for short FMS; The Red5 servers of increasing income etc. mainly are applied to online live, recorded video aspect.For this agreement, need to revise some application and connect pairing server end ASC this opening of angle client authority, allow the Flash playing program of client can preserve data to internal memory.Specific code is following:
client.videoSampleAccess=″/″;
Certainly, also can dispose as follows through directly in the configuration file XML of server, overall authority being set:
<VideoSampleAccess?enabled=″true″>/</VideoSampleAccess>
It is as shown in Figure 3 to the process flow diagram of the authority of internal memory that online video data is preserved in above-mentioned acquisition.
Embodiment 6: as an application to above-mentioned each embodiment; In the present embodiment the single frames data after the recognition of face are carried out application processes; For example; Just can extract position coordinates, the size at people's face place according to face recognition result, on its basis, realize all kinds of application to the dynamic video picture.Comprise that people's face of each dvielement is followed effect, human face region special efficacy etc.
Embodiment 7: corresponding with the foregoing description 1 to embodiment 6; Corresponding face identification device is provided in the present embodiment; As shown in Figure 4, comprise processing module, system interface and face recognition module, principle of work is following: the real resource utilization rate that processing module computing client end is current; Regulate identification parameter according to said current real resource utilization rate; System interface is sampled to said Online Video data according to the identification parameter after regulating; Face recognition module is carried out recognition of face to current Online Video data of sampling.In addition, to embodiment 7, face identification device can also comprise master controller, and face recognition module returns to master controller with recognition result, and master controller carries out application processes to the single frames picture.
The foregoing description is a preferred implementation of the present invention; But embodiment of the present invention is not restricted to the described embodiments; Other are any not to deviate from modification, the modification done under spirit of the present invention and the principle, substitute, combination, simplify; All should be the substitute mode of equivalence, all should be included within protection scope of the present invention.
Claims (10)
1. the face identification method of a Flash Online Video is characterized in that, comprises step:
The real resource utilization rate that the computing client end is current;
Regulate identification parameter according to said current real resource utilization rate, said Online Video data are sampled according to the identification parameter after regulating;
Through face recognition module current Online Video data of sampling are carried out recognition of face.
2. face identification method according to claim 1 is characterized in that, said identification parameter comprises the image sampling size of said Online Video data.
3. face identification method according to claim 1 and 2 is characterized in that, said identification parameter comprises the image sampling frequency of said Online Video data.
4. face identification method according to claim 3 is characterized in that, current Online Video data of sampling is carried out recognition of face also comprise step afterwards:
Extract the position coordinates and the size at people's face place, and carry out application processes, said application processes comprise following one or more: people's face of human face region special efficacy, each dvielement is followed effect.
5. according to the face identification method of claim 1, it is characterized in that the step that said Online Video data are sampled also comprises before:
If said Online Video data then load the cross-domain strategy file in the pairing server of said Online Video data through the http protocol transmission, to obtain to preserve the authority of this video data;
If said Online Video data through the rmpt protocol transmission, obtain to preserve through following steps the authority of this video data: client.videoSampleAccess="/" or
Configuration file in the pairing server of said Online Video data is provided with overall authority, to obtain to preserve the authority of this video data.
6. the face identification device of a Flash Online Video is characterized in that, comprising:
Processing module is used for the current real resource utilization rate of computing client end; Regulate identification parameter according to said current real resource utilization rate;
System interface is used for according to the identification parameter after regulating said Online Video data being sampled;
Face recognition module is carried out recognition of face to current Online Video data of sampling.
7. face identification device according to claim 6 is characterized in that, said identification parameter comprises the image sampling size of said Online Video data.
8. according to claim 6 or 7 described face identification devices, it is characterized in that said identification parameter comprises the image sampling frequency of said Online Video data.
9. face identification device according to claim 8; It is characterized in that; Also comprise master controller; Be used to extract the position coordinates and the size at people's face place, the current image that imports into carried out application processes, said application processes comprise following one or more: people's face of human face region special efficacy, each dvielement is followed effect.
10. according to the face identification device of claim 6; It is characterized in that; If said Online Video data are through the http protocol transmission, system interface loads the strategy file in the pairing server of said Online Video data, to obtain to preserve the authority of this video data.
If said Online Video data are through the rmpt protocol transmission, system interface obtains to preserve the authority of this video data: client.videoSampleAccess="/" through following steps.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2011104447218A CN102567720A (en) | 2011-12-26 | 2011-12-26 | Face identification method and face identification device for Flash online video |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2011104447218A CN102567720A (en) | 2011-12-26 | 2011-12-26 | Face identification method and face identification device for Flash online video |
Publications (1)
Publication Number | Publication Date |
---|---|
CN102567720A true CN102567720A (en) | 2012-07-11 |
Family
ID=46413097
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2011104447218A Pending CN102567720A (en) | 2011-12-26 | 2011-12-26 | Face identification method and face identification device for Flash online video |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102567720A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108279855A (en) * | 2015-03-05 | 2018-07-13 | 青岛海信宽带多媒体技术有限公司 | A method of read-write storage device |
CN108304753A (en) * | 2017-01-24 | 2018-07-20 | 腾讯科技(深圳)有限公司 | Video communication method and video communication device |
CN109302477A (en) * | 2018-09-30 | 2019-02-01 | 武汉斗鱼网络科技有限公司 | A kind of dispatching method and relevant apparatus of task |
CN110035098A (en) * | 2018-01-12 | 2019-07-19 | 上海杰轩智能科技有限公司 | The method for realizing recognition of face terminal data transmission by XML format |
CN111862617A (en) * | 2020-06-12 | 2020-10-30 | 浙江大华技术股份有限公司 | License plate recognition method, device and system and computer equipment |
WO2021159672A1 (en) * | 2020-02-11 | 2021-08-19 | 深圳壹账通智能科技有限公司 | Face image recognition method and apparatus |
CN113537090A (en) * | 2021-07-21 | 2021-10-22 | 南京奥拓电子科技有限公司 | Intelligent household lamp control method and device and storage medium |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101325054A (en) * | 2007-06-13 | 2008-12-17 | 三星电子株式会社 | Display system and method for displaying video signals |
US20090290789A1 (en) * | 2008-05-23 | 2009-11-26 | Advanced Micro Devices, Inc. | Method and apparatus for reduced complexity video processing via special chroma handling |
-
2011
- 2011-12-26 CN CN2011104447218A patent/CN102567720A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101325054A (en) * | 2007-06-13 | 2008-12-17 | 三星电子株式会社 | Display system and method for displaying video signals |
US20090290789A1 (en) * | 2008-05-23 | 2009-11-26 | Advanced Micro Devices, Inc. | Method and apparatus for reduced complexity video processing via special chroma handling |
Non-Patent Citations (1)
Title |
---|
何柯峰: "在线人脸识别***", 《合肥工业大学硕士学位论文》 * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108279855A (en) * | 2015-03-05 | 2018-07-13 | 青岛海信宽带多媒体技术有限公司 | A method of read-write storage device |
CN108304753A (en) * | 2017-01-24 | 2018-07-20 | 腾讯科技(深圳)有限公司 | Video communication method and video communication device |
CN110035098A (en) * | 2018-01-12 | 2019-07-19 | 上海杰轩智能科技有限公司 | The method for realizing recognition of face terminal data transmission by XML format |
CN109302477A (en) * | 2018-09-30 | 2019-02-01 | 武汉斗鱼网络科技有限公司 | A kind of dispatching method and relevant apparatus of task |
WO2021159672A1 (en) * | 2020-02-11 | 2021-08-19 | 深圳壹账通智能科技有限公司 | Face image recognition method and apparatus |
CN111862617A (en) * | 2020-06-12 | 2020-10-30 | 浙江大华技术股份有限公司 | License plate recognition method, device and system and computer equipment |
CN113537090A (en) * | 2021-07-21 | 2021-10-22 | 南京奥拓电子科技有限公司 | Intelligent household lamp control method and device and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102567720A (en) | Face identification method and face identification device for Flash online video | |
CN111901674B (en) | Video playing control method and device | |
CN105637887B (en) | Method for video impression analysis | |
US10313726B2 (en) | Distributing media content via media channels based on associated content being provided over other media channels | |
US9226005B2 (en) | Preprocessing video to insert visual elements and applications thereof | |
US20220007083A1 (en) | Method and stream-pushing client for processing live stream in webrtc | |
US8928810B2 (en) | System for combining video data streams into a composite video data stream | |
CN111064973A (en) | Live broadcast system based on IPV9 | |
US20140123162A1 (en) | Eye tracking based defocusing | |
US11563991B2 (en) | Methods and apparatus to monitor a media presentation | |
US20160295269A1 (en) | Information pushing method, device and system | |
CN111064972A (en) | Live video control method based on IPV9 | |
CN103096182A (en) | Network television program information sharing method and system | |
JP2014032669A5 (en) | ||
US10972511B2 (en) | Streaming relay for digital signage | |
CN113535063A (en) | Live broadcast page switching method, video page switching method, electronic device and storage medium | |
CN106454388A (en) | Method and device for determining live broadcast setting information | |
US20220182732A1 (en) | Systems and Methods of Alternative Networked Application Services | |
CN114139491A (en) | Data processing method, device and storage medium | |
CN112492324A (en) | Data processing method and system | |
US20230108310A1 (en) | Live video latency measurement and control | |
EP3334171A1 (en) | System for providing cloud-based user interfaces and method thereof | |
CN113068059B (en) | Video live broadcasting method, device, equipment and storage medium | |
CN111787417B (en) | Audio and video transmission control method based on artificial intelligence AI and related equipment | |
WO2015042767A1 (en) | Method for inserting an advertisement into a video stream of an application on demand (aod) service, aod processing device and aod server |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C12 | Rejection of a patent application after its publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20120711 |