CN104636736A - High-definition face detecting method based on libjpeg base - Google Patents
High-definition face detecting method based on libjpeg base Download PDFInfo
- Publication number
- CN104636736A CN104636736A CN201510094072.1A CN201510094072A CN104636736A CN 104636736 A CN104636736 A CN 104636736A CN 201510094072 A CN201510094072 A CN 201510094072A CN 104636736 A CN104636736 A CN 104636736A
- Authority
- CN
- China
- Prior art keywords
- data
- face
- libjpeg
- internal storage
- jpeg
- 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
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
- G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a high-definition face detecting method based on a libjpeg base. The high-definition face detecting method comprises the following steps: 1) obtaining compressed face-containing JPEG memory data from an image data interface of a front-end high definition camera through thread control; 2) decoding the original face-containing JPEG memory data through the libjpeg base; 3) performing 16-byte aligning processing on the decoded face-containing image data in the memory unit, and extracting face data; 4) getting out the face data, converting the face data into YUV data, and converting the YUV data into an RGB24-format image; 5) storing or sending. The high-definition face detecting method adopts an open high-performance high-reliability uncompressing method to realize memory data decoding, and can perform any image processing on decoded data according to the needs, so that the extraction, the display, the re-compression, the storage and the network transmission of the face are realized, the effective data loss of the face is reduced to the greatest extent, the face distortion is reduced, the face definition is guaranteed, and the intelligent level of a security and protection system is improved.
Description
Technical field
The present invention discloses a kind of high definition method for detecting human face based on libjpeg storehouse.
Background technology
Along with high definition, networking, digitized development, safety defense monitoring system obtains a large amount of universal installation, and video monitoring data increases with surprising rapidity.
At present, in many banks, company, the place such as garden, public place is all provided with the video monitoring of 7 × 24 hours.When having abnormal conditions or have stranger to swarm into, need real-time follow-up, monitoring, identification and warning etc.These needs are made a concrete analysis of the image collected, and will use the detection of face, follow the tracks of and recognition technology.
But the method for current coding/decoding method and image procossing, the loss of the face valid data that can cause, distortion, there will be the problems such as face sharpness owing to not supporting high definition decoding etc. to cause is inadequate simultaneously.
Summary of the invention
The object of the invention is to overcome the defect existed in prior art, a kind of loss reducing face valid data is to greatest extent provided, reduce the distortion of face, ensure the sharpness of face, improve the high definition method for detecting human face based on libjpeg storehouse of the intelligent level of safety-protection system.
The present invention is achieved in that a kind of high definition method for detecting human face based on libjpeg storehouse, it is characterized in that: comprise the steps:
1) the JPEG internal storage data containing face is obtained: the JPEG internal storage data containing face being obtained compression by Thread control from the image data interface of front end high-definition camera; Wherein said containing face JPEG internal storage data comprises figure image height data, figure image width data, view data size, face coordinate figure, face is high, face is wide, face quantity;
2) the described JPEG internal storage data containing face is decoded: by libjpeg storehouse, data input is carried out as decoding parametric using described figure image height data, figure image width data, view data size, decode containing the JPEG internal storage data of face described in obtaining, obtain the decoded view data containing face, then the described view data containing face is exported and be stored in internal storage location;
3) registration process extract human face data: to the decoded view data containing face in described internal storage location after the registration process through 16 bytes, according to described face coordinate figure, extract human face data line by line, the human face data of extraction is stored in internal storage location for step 4;
4) format transformation: take out the human face data stored in internal storage location in step 3, and convert yuv data to, and then convert RGB24 format-pattern to;
5) store or send: the RGB24 format-pattern in step 4 stored by external memory unit or is sent to central platform by communication module.
After described step 5, also comprise step 6: described RGB24 format-pattern is compressed into jpg file by described libjpeg storehouse by described central platform, and store or show jpg file by display unit.
In described step 2, the described decoded view data size containing face is figure image height data × figure image width data described in the step 1 of 3 times.
In described step 2, carrying out in decode procedure to the described JPEG internal storage data containing face, comprise: carry out data structure definition to the described JPEG internal storage data containing face, the compressing information comprising definition jpeg data is for as calling the parameter of libjpeg bank interface, the error message of definition jpeg data for other information of catching, defining jpeg data of mistake in decode procedure temporal cache for decoding a line view data; After completing data structure definition, carry out decoding initialization, and specify decompressed data source, obtain containing the JPEG internal storage data of face, start decompress(ion), calculate the often memory size that needs of row of the picture after decompress(ion), then decode line by line; The decoded view data containing face exports and is stored in internal storage location.
Communication module in described step 5 is socket communication module.
Described RGB24 format-pattern is compressed into jpg file by described libjpeg storehouse and comprises step by described central platform:
The RGB24 form data structure used of encoding is defined, and bind standard error process structure, then just decoding beginningization is carried out, intended target image file, arranges acquiescence compression parameters, arranges coding quality, starts coding, carries out encoding, having encoded and saved as jpg file line by line.
The invention has the beneficial effects as follows: support of the present invention obtains the jpeg data containing face of compression from the high-definition camera of high definition 720P or 1080P, the highly reliable decompression method of open high-performance is adopted to realize internal storage data decoding, arbitrary image procossing can be carried out as required to decoded data, to realize the extraction of face, display, recompression, storage and Internet Transmission, decrease the loss of face valid data to greatest extent, reduce the distortion of face, ensure that the sharpness of face, improve the intelligent level of safety-protection system.
Accompanying drawing explanation
Fig. 1 is method flow diagram of the present invention.
Fig. 2 is that the present invention is to described process flow diagram of decoding containing the JPEG internal storage data of face.
Fig. 3 is that described RGB24 format-pattern is compressed into the process flow diagram of jpg file by the present invention.
Embodiment
According to Fig. 1 ~ Fig. 3, a kind of high definition method for detecting human face based on libjpeg storehouse of the present invention: comprise the steps:
1) the JPEG internal storage data containing face is obtained: the JPEG internal storage data containing face being obtained compression by Thread control from the image data interface of front end high-definition camera; Wherein said containing face JPEG internal storage data comprises figure image height data, figure image width data, view data size, face coordinate figure, face is high, face is wide, face quantity.
2) the described JPEG internal storage data containing face is decoded: according to Fig. 2, by libjpeg storehouse, carry out data input using described figure image height data, figure image width data, view data size as decoding parametric; Then data structure definition is carried out to the described JPEG internal storage data containing face: the compressing information comprising definition jpeg data is for as calling the parameter of libjpeg bank interface, the error message of definition jpeg data for other information of catching, defining jpeg data of mistake in decode procedure temporal cache for decoding a line view data.After completing data structure definition, carry out decoding initialization, specify decompressed data source, obtain containing the JPEG internal storage data of face, start decompress(ion), calculate the often memory size that needs of row of the picture after decompress(ion), then decode line by line; The decoded view data containing face exports and is stored in internal storage location, terminates decoding after release decode resources.The wherein said decoded view data size containing face is figure image height data × figure image width data described in the step 1 of 3 times.
3) registration process extract human face data: to the decoded view data containing face in described internal storage location after the registration process through 16 bytes, according to described face coordinate figure, extract human face data line by line, the human face data of extraction is stored in internal storage location for step 4;
4) format transformation: take out the human face data stored in internal storage location in step 3, and convert yuv data to, and then convert RGB24 format-pattern to.
5) store or send: the RGB24 format-pattern in step 4 stored by external memory unit or is sent to central platform by communication module; Described communication module is socket communication module.
6) recompress: according to Fig. 3, described RGB24 format-pattern is compressed into jpg file by described libjpeg storehouse by described central platform, and store or show jpg file by display unit.
Described RGB24 format-pattern is compressed into jpg file by described libjpeg storehouse and comprises step by described central platform: define the RGB24 form data structure used of encoding, and bind standard error process structure, then coding initialization is carried out, intended target image file, acquiescence compression parameters is set, coding quality is set, starts coding, carries out encoding, having encoded and saved as jpg file line by line, after releasing resource, terminate decoding.
The present invention, by obtaining the JPEG internal storage data containing face of high definition compression, adopts the highly reliable decompression algorithm of open high-performance to realize internal storage data decoding, ensure that the sharpness of face; Simultaneously to the algorithm of decoded image real time transfer devise optimum, realize the extraction of face, display, recompression, storage and Internet Transmission, decrease the loss of face valid data to greatest extent, reduce the distortion of face.
Claims (6)
1., based on the high definition method for detecting human face in libjpeg storehouse, it is characterized in that: comprise the steps:
1) the JPEG internal storage data containing face is obtained: the JPEG internal storage data containing face being obtained compression by Thread control from the image data interface of front end high-definition camera; Wherein said containing face JPEG internal storage data comprises figure image height data, figure image width data, view data size, face coordinate figure, face is high, face is wide, face quantity;
2) the described JPEG internal storage data containing face is decoded: by libjpeg storehouse, data input is carried out as decoding parametric using described figure image height data, figure image width data, view data size, decode containing the JPEG internal storage data of face described in obtaining, obtain the decoded view data containing face, then the described view data containing face is exported and be stored in internal storage location;
3) registration process extract human face data: to the decoded view data containing face in described internal storage location after the registration process through 16 bytes, according to described face coordinate figure, extract human face data line by line, the human face data of extraction is stored in internal storage location for step 4;
4) format transformation: take out the human face data stored in internal storage location in step 3, and convert yuv data to, and then convert RGB24 format-pattern to;
5) store or send: the RGB24 format-pattern in step 4 stored by external memory unit or is sent to central platform by communication module.
2. according to a kind of high definition method for detecting human face based on libjpeg storehouse described in claim 1, it is characterized in that: after described step 5, also comprise step 6: described RGB24 format-pattern is compressed into jpg file by described libjpeg storehouse by described central platform, and store or show jpg file by display unit.
3. according to a kind of high definition method for detecting human face based on libjpeg storehouse described in claim 1, it is characterized in that: in described step 2, the described decoded view data size containing face is figure image height data × figure image width data described in the step 1 of 3 times.
4. according to a kind of high definition method for detecting human face based on libjpeg storehouse described in claim 1, it is characterized in that: in described step 2, carrying out in decode procedure to the described JPEG internal storage data containing face, comprise: carry out data structure definition to the described JPEG internal storage data containing face, the compressing information comprising definition jpeg data is for as calling the parameter of libjpeg bank interface, the error message of definition jpeg data for other information of catching, defining jpeg data of mistake in decode procedure temporal cache for decoding a line view data; After completing data structure definition, carry out decoding initialization, and specify decompressed data source, obtain containing the JPEG internal storage data of face, start decompress(ion), calculate the often memory size that needs of row of the picture after decompress(ion), then decode line by line; The decoded view data containing face exports and is stored in internal storage location.
5., according to a kind of high definition method for detecting human face based on libjpeg storehouse described in claim 1, it is characterized in that: the communication module in described step 5 is socket communication module.
6. according to a kind of high definition method for detecting human face based on libjpeg storehouse described in claim 2, it is characterized in that: described RGB24 format-pattern is compressed into jpg file by described libjpeg storehouse and comprises step by described central platform: the RGB24 form data structure used of encoding is defined, and bind standard error process structure, then decoding initialization is carried out, intended target image file, arranges acquiescence compression parameters, arranges coding quality, starts coding, carries out encoding, having encoded and saved as jpg file line by line.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510094072.1A CN104636736A (en) | 2015-03-03 | 2015-03-03 | High-definition face detecting method based on libjpeg base |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510094072.1A CN104636736A (en) | 2015-03-03 | 2015-03-03 | High-definition face detecting method based on libjpeg base |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104636736A true CN104636736A (en) | 2015-05-20 |
Family
ID=53215466
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510094072.1A Pending CN104636736A (en) | 2015-03-03 | 2015-03-03 | High-definition face detecting method based on libjpeg base |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104636736A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106611149A (en) * | 2015-10-22 | 2017-05-03 | 腾讯科技（深圳）有限公司 | Method and device for positioning skin area |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1928922A (en) * | 2006-09-22 | 2007-03-14 | 浙江大学 | Image processing method for face ageing |
| WO2009062945A1 (en) * | 2007-11-16 | 2009-05-22 | Seereal Technologies S.A. | Method and device for finding and tracking pairs of eyes |
-
2015
- 2015-03-03 CN CN201510094072.1A patent/CN104636736A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1928922A (en) * | 2006-09-22 | 2007-03-14 | 浙江大学 | Image processing method for face ageing |
| WO2009062945A1 (en) * | 2007-11-16 | 2009-05-22 | Seereal Technologies S.A. | Method and device for finding and tracking pairs of eyes |
Non-Patent Citations (1)
| Title |
|---|
| 李景坤 等: "《家庭视觉智能防盗监控***》", 《微型机与应用》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106611149A (en) * | 2015-10-22 | 2017-05-03 | 腾讯科技（深圳）有限公司 | Method and device for positioning skin area |
| CN106611149B (en) * | 2015-10-22 | 2019-12-13 | 腾讯科技（深圳）有限公司 | method and device for locating skin area |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6338688B2 (en) | Video synchronized playback method, apparatus, and system | |
| WO2018150083A1 (en) | A method and technical equipment for video processing | |
| EP2733939A1 (en) | Encoding/decoding method and apparatus | |
| CN107730759A (en) | A kind of intelligent express delivery cabinet pickup system based on face recognition technology | |
| US20220130151A1 (en) | Surveillance systems and methods | |
| CN104320622A (en) | Embedded video enhancement system for open source server software | |
| CN105554513A (en) | Panoramic video transmission method and system based on H.264 | |
| CN107172433A (en) | A kind of video encryption method | |
| WO2021057684A1 (en) | Video decoding method and apparatus, video encoding method and apparatus, storage medium, and electronic apparatus | |
| WO2023045364A1 (en) | Image display method and apparatus, and storage medium and electronic apparatus | |
| CN112584155A (en) | Video data processing method and device | |
| CN104636736A (en) | High-definition face detecting method based on libjpeg base | |
| CN113747191A (en) | Video live broadcast method, system, equipment and storage medium based on unmanned aerial vehicle | |
| CN103561264A (en) | Media decoding method based on cloud computing and decoder | |
| CN104244085B (en) | Multimedia data transmission method and device based on field programmable gate array | |
| CN111010534A (en) | Real-time asynchronous video analysis method and system | |
| CN107071455A (en) | Jpeg image information concealing method based on data flow | |
| CN105578008A (en) | Method of improving USB 2.0 camera high resolution high frame rate based on android terminal | |
| CN105872552A (en) | Intelligent processing method for prolonging video storage time of video monitoring system | |
| WO2022110870A1 (en) | Image encoding and decoding method, encoding and decoding apparatus, encoder, and decoder | |
| CN112954456B (en) | Video data processing method, terminal and computer readable storage medium | |
| CN105578143A (en) | Intelligent remote monitoring system | |
| CN114640882B (en) | Video processing method, video processing device, electronic equipment and computer readable storage medium | |
| CN108508799A (en) | A kind of long-range control method | |
| WO2020107376A1 (en) | Image processing method, device, and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150520 |