CN110765919A - Interview image display system and method based on face detection - Google Patents
Interview image display system and method based on face detection Download PDFInfo
- Publication number
- CN110765919A CN110765919A CN201910992634.2A CN201910992634A CN110765919A CN 110765919 A CN110765919 A CN 110765919A CN 201910992634 A CN201910992634 A CN 201910992634A CN 110765919 A CN110765919 A CN 110765919A
- Authority
- CN
- China
- Prior art keywords
- image
- face
- face detection
- detection
- interview
- 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
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
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/94—Hardware or software architectures specially adapted for image or video understanding
-
- 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
-
- 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
- G06V40/172—Classification, e.g. identification
Abstract
The invention discloses a interview image display system and method based on face detection, and belongs to the technical field of computer image processing. The visiting image display system based on the face detection comprises a plurality of network cameras, an image processing host, a screen display end and a visiting interaction end, wherein the image processing host comprises an image acquisition unit, a face detection unit, a face feature database, an image analysis unit and an image storage unit; the screen display end comprises a screen display module and a display screen, and the image storage unit, the screen display module and the display screen are sequentially connected. The interview image display system based on the face detection can enhance the user experience of interviewers and has good popularization and application values.
Description
Technical Field
The invention relates to the technical field of computer image processing, and particularly provides a interview image display system and method based on face detection.
Background
The enterprise exhibition hall is taken as an important place for displaying products and culture of enterprises, and has attracted more and more attention of the enterprises, and the construction of the exhibition hall is developing towards intellectualization. The intelligent exhibition hall aims to enable exhibition to be more intelligent and modern, specifically realizes centralized control of various exhibition items, provides multimedia audio-visual environment control, increases interactive experience of interviewer, and is also one aspect of automatic analysis and exhibition of interview images.
Typically, the corporate leader will take the interview to visit and introduce the product, and multiple parties will visit in sequence. After the whole visiting process is completed, the enterprise hopes to record the visiting process in multiple visual angles, and intelligently extracts the most valuable images, so that the visiting personnel can check and store the photos of the group of the enterprise in time, and the process does not need the intervention of a guide person, thereby increasing the interactive experience of the visiting personnel and the enterprise, and improving the intelligent level of an enterprise exhibition hall.
Disclosure of Invention
The technical task of the invention is to provide a interview image display system based on face detection, which can check and store group photos without adding auxiliary equipment and can enhance the user experience of interviewers.
The invention further provides a interview image display method based on face detection.
In order to achieve the purpose, the invention provides the following technical scheme:
a interview image display system based on face detection comprises a plurality of network cameras, an image processing host, a screen display end and an interview interaction end, wherein the image processing host comprises an image acquisition unit, a face detection unit, a face feature database, an image analysis unit and an image storage unit; the screen display end comprises a screen display module and a display screen, and the image storage unit, the screen display module and the display screen are sequentially connected; the interview interaction terminal comprises an interview interaction module and a mobile equipment display module, and the image storage unit, the interview interaction module and the mobile equipment display module are sequentially connected.
Preferably, the image processing host further comprises a GPU acceleration module, and the GPU acceleration module is respectively connected with the face detection unit and the image analysis unit.
And the face detection unit calls a GPU acceleration module to perform face detection on the image. And the image analysis unit calls a GPU acceleration module to analyze the image.
Preferably, the network cameras support a real-time streaming protocol, and the image acquisition unit is connected with each network camera through the real-time streaming protocol.
Preferably, the face detection unit performs image face detection based on a convolutional neural network, and stores the detected image face into a face feature database.
Preferably, the face detection unit is further configured to update the shooting states of the network cameras, and mark that a current camera is in a detection state, a previous camera is in a detection stop state, and a next adjacent camera is to be detected after a certain camera detects the face features in two consecutive detection periods.
Preferably, the content stored in the face feature database comprises the number of detected faces, each face feature vector, the pixel size of each face and the number of the detected network camera; the face feature database stores network camera information, preset key figure face feature vectors and excluded face feature vectors in advance.
A visiting image display method based on face detection is realized based on the visiting image display system based on face detection, a plurality of network cameras are arranged in multiple areas of an exhibition hall to obtain real-time images in the exhibition hall, an image acquisition unit reads real-time video streams of the network cameras and extracts the video streams into one image to be transmitted to a face detection unit, the face detection unit carries out image face detection and stores face features into a face feature database, an image analysis unit carries out analysis according to the face features in the face feature database and stores the analyzed face image into an image storage unit, and images are viewed through a display screen and a mobile equipment display end.
Preferably, the face detection unit calls the GPU acceleration module to perform face detection on the image based on the convolutional neural network, when the front face features are not detected, the detection is repeated, and when the face features are detected, the number of front faces, the pixel size of each front face and each face feature vector contained in the image are calculated.
Preferably, the face detection unit is configured to update the shooting states of the respective network cameras, and mark that a current camera is in a detection state, a previous camera is in a detection stop state, and a next adjacent camera is to be detected after a certain camera detects face features in two consecutive detection periods.
Preferably, when the number of faces contained in the detected image is lower than the group minimum number or the size of the front face pixel is smaller than the size of the face of the normal visitor or compared with a face feature excluding library preset in a face feature database, if the image contains face features, the image is an invalid image, otherwise, the image is a normal valid image.
Compared with the prior art, the interview image display system based on the face detection has the following outstanding beneficial effects: the interview image display system based on the face detection is characterized in that the image acquisition unit acquires video streams of the network camera in real time, the face detection unit performs face feature detection on the video images, the face feature detection can be matched with various preset conditions, different interview bodies are distinguished through face recognition clustering, various image evaluation rules can be set for image optimization, the optimized images can be checked on a display screen and a mobile device, a leader auxiliary device is not required to be added in the whole process, the user experience of interviewers is enhanced, and the interview image display system has good popularization and application values.
Drawings
Fig. 1 is a frame diagram of a interview image presentation system based on face detection according to the present invention.
Detailed Description
The following describes the face detection-based interview image presentation system and method in further detail with reference to the accompanying drawings and examples.
Examples
As shown in fig. 1, the interview image display system based on face detection of the present invention includes a plurality of network cameras, an image processing host, a screen display end and an interview interaction end.
The image processing host comprises an image acquisition unit, a face detection unit, a face feature database, an image analysis unit, an image storage unit and a GPU acceleration module.
The plurality of network cameras are respectively arranged at each corner of the exhibition hall, the plurality of network cameras support a real-time stream transmission protocol, and the image acquisition unit is connected with each network camera through the real-time stream transmission protocol. The image acquisition unit, the face detection unit, the face feature database, the image analysis unit and the image storage unit are sequentially connected. The GPU acceleration module is respectively connected with the face detection unit and the image analysis unit.
The face detection unit calls the GPU acceleration module to perform face detection on the image, the face detection unit performs image face detection on the basis of the convolutional neural network, and the detected image face is stored in a face feature database. The face detection unit is also used for updating the shooting state of each network camera, when a certain camera detects face features in two continuous detection periods, the current camera is marked to be in a detection state, the previous camera is marked to be in a detection stop state, and the next adjacent camera is subjected to detection.
And the image analysis unit calls the GPU acceleration module to analyze the image.
The contents stored in the face feature database comprise the number of detected faces, each face feature vector, the pixel size of each face and the serial number of a detected network camera; the face feature database stores network camera information, preset key figure face feature vectors and excluded face feature vectors in advance.
The screen display end comprises a screen display module and a display screen, and the image storage unit, the screen display module and the display screen are sequentially connected.
The interview interaction terminal comprises an interview interaction module and a mobile equipment display module, and the image storage unit, the interview interaction module and the mobile equipment display module are sequentially connected.
The visiting image display method based on the face detection is realized based on the visiting image display system based on the face detection.
The method comprises the steps that a plurality of network cameras are arranged in multiple areas of an exhibition hall, real-time images in the exhibition hall are obtained, an image acquisition unit reads real-time video streams of the network cameras, the video streams are extracted into one image and transmitted to a face detection unit, the face detection unit carries out image face detection and stores face features into a face feature database, an image analysis unit carries out analysis according to the face features in the face feature database and stores the analyzed face images into an image storage unit, and the images are checked through a display screen and a mobile equipment display end.
And the face detection unit calls a GPU acceleration module to perform face detection on the image based on the convolutional neural network, when the face features are not detected, the detection is repeated, and when the face features are detected, the number of the face faces on the front side, the pixel size of each face on the front side and the feature vector of each face contained in the image are calculated. The face detection unit is used for updating the shooting state of each network camera, when a certain camera detects face features in two continuous detection periods, the current camera is marked to be in a detection state, the previous camera is marked to be in a detection stop state, and the next adjacent camera is subjected to detection.
And when the number of the faces contained in the detected image is lower than the minimum group number or the size of the front face pixels is smaller than the size of the faces of normal interviewer or compared with a face feature exclusion library preset in a face feature database, if the image contains the face features, the image is an invalid image, and if not, the image is a normal valid image.
After the current network camera detects an effective image, the face detection unit marks the current network camera as a detection state, and creates a grouping identifier for the current group for distinguishing.
And updating the adjacent last numbered network camera to be in a detection ending state, and processing the condition that team personnel are simultaneously distributed in two areas in the advancing process.
And after the continuous invalid times of the detection images reach a set threshold value, ending the group detection and marking that the grouping of the areas is ended.
And when the network camera in the last area finishes the grouping detection, all the grouping images which are not optimized are transmitted to the image analysis unit.
The image analysis unit further groups and optimizes the images containing the human face features, and comprises the following processes:
1) and extracting the facial features in the groups and between adjacent groups according to the primary grouping result of the images by the facial detection unit, performing clustering comparison again, merging the two groups of images if the facial features of the images between the two groups belong to the same group, and enabling each group of processed images to only contain the same group in the same region.
2) Preferably, each group of images is screened out a certain number of images as the best image of a region by using the following rules:
the number of the front faces contained in the multiple images in the same group is large;
matching the optimal face size of the corresponding shooting area;
and matching preset key figure face characteristics.
3) And combining the images of the same community in different areas to form a new group, wherein the new group represents the image of the community in the whole visiting process.
4) And outputting the image to a screen display end.
The screen display end comprises a screen display module and a display screen, the latest group of image output value display screens are displayed in turn, and the group interview interaction end access two-dimensional code is generated.
The visiting interaction terminal provides a display page which is accessed by the mobile equipment such as a mobile phone through scanning the two-dimensional code, the page displays all regional photos of the current group in the visiting process, and the visiting personnel can store the original pictures to the mobile phone. And when the two-dimension code exceeds the effective period, the link is accessed again, and the reference image is not checked.
The above-described embodiments are merely preferred embodiments of the present invention, and general changes and substitutions by those skilled in the art within the technical scope of the present invention are included in the protection scope of the present invention.
Claims (10)
1. A interview image display system based on face detection is characterized in that: the system comprises a plurality of network cameras, an image processing host, a screen display end and a visiting interaction end, wherein the image processing host comprises an image acquisition unit, a face detection unit, a face feature database, an image analysis unit and an image storage unit; the screen display end comprises a screen display module and a display screen, and the image storage unit, the screen display module and the display screen are sequentially connected; the interview interaction terminal comprises an interview interaction module and a mobile equipment display module, and the image storage unit, the interview interaction module and the mobile equipment display module are sequentially connected.
2. The face detection-based interview image presentation system of claim 1, wherein: the image processing host further comprises a GPU acceleration module, and the GPU acceleration module is respectively connected with the face detection unit and the image analysis unit.
3. The face detection-based interview image presentation system of claim 2, wherein: the plurality of network cameras support a real-time streaming protocol, and the image acquisition unit is connected with each network camera through the real-time streaming protocol.
4. The face detection-based interview image presentation system of claim 3, wherein: the face detection unit performs image face detection based on the convolutional neural network and stores the detected image face into a face feature database.
5. The face detection-based interview image presentation system of claim 4, wherein: the face detection unit is also used for updating the shooting state of each network camera, when a certain camera detects face features in two continuous detection periods, the current camera is marked to be in a detection state, the previous camera is marked to be in a detection stop state, and the next adjacent camera is subjected to detection and the like.
6. The face detection-based interview image presentation system of claim 5, wherein: the contents stored in the face feature database comprise the number of detected faces, each face feature vector, the pixel size of each face and the serial number of a detected network camera; the face feature database stores network camera information, preset key figure face feature vectors and excluded face feature vectors in advance.
7. A interview image display method based on face detection is characterized in that: the method is realized based on the visiting image display system based on the face detection as claimed in any one of claims 1 to 6, a plurality of network cameras are arranged in a plurality of areas of an exhibition hall to obtain real-time images in the exhibition hall, an image acquisition unit reads real-time video streams of the network cameras and extracts the video streams into one image to be transmitted to a face detection unit, the face detection unit carries out image face detection, face features are stored in a face feature database, an image analysis unit carries out analysis according to the face features in the face feature database, the analyzed face image is stored in an image storage unit, and the image is viewed through a display screen and a mobile equipment display terminal.
8. The method for displaying a visiting image based on face detection as claimed in claim 7, wherein: the face detection unit calls a GPU acceleration module to perform face detection on the image based on a convolutional neural network, when the face features are not detected, the face detection unit repeatedly detects the face features, and when the face features are detected, the number of the face faces on the front side, the pixel size of each face on the front side and the feature vector of each face contained in the image are calculated.
9. The method for displaying a visiting image based on face detection according to claim 8, wherein: the face detection unit is used for updating the shooting state of each network camera, when a certain camera detects face features in two continuous detection periods, the current camera is marked to be in a detection state, the previous camera is marked to be in a detection stop state, and the next adjacent camera is subjected to detection.
10. The method for displaying a visiting image based on face detection according to claim 9, wherein: and when the number of the faces contained in the detected image is lower than the minimum group number or the size of the front face pixels is smaller than the size of the faces of normal interviewer or compared with a face feature exclusion library preset in a face feature database, if the image contains the face features, the image is an invalid image, and if not, the image is a normal valid image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910992634.2A CN110765919A (en) | 2019-10-18 | 2019-10-18 | Interview image display system and method based on face detection |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910992634.2A CN110765919A (en) | 2019-10-18 | 2019-10-18 | Interview image display system and method based on face detection |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110765919A true CN110765919A (en) | 2020-02-07 |
Family
ID=69332427
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910992634.2A Pending CN110765919A (en) | 2019-10-18 | 2019-10-18 | Interview image display system and method based on face detection |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110765919A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113536038A (en) * | 2021-08-03 | 2021-10-22 | 深圳市一朴创意有限责任公司 | Intelligent exhibition hall distributed control method, device, system, equipment and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102622580A (en) * | 2012-02-20 | 2012-08-01 | 华焦宝 | Face detection and recognition method and system |
CN102916931A (en) * | 2011-08-03 | 2013-02-06 | 北京同步科技有限公司 | Multimedia information distributing system and method |
US20170193772A1 (en) * | 2015-12-31 | 2017-07-06 | Cerner Innovation, Inc. | Detecting unauthorized visitors |
CN108875695A (en) * | 2018-07-05 | 2018-11-23 | 深圳勇艺达机器人有限公司 | Visit scene interest stream of people's statistical analysis technique based on machine vision |
CN110298706A (en) * | 2019-07-04 | 2019-10-01 | 山东浪潮人工智能研究院有限公司 | A kind of exhibition room visits interaction and takes pictures evaluation method |
-
2019
- 2019-10-18 CN CN201910992634.2A patent/CN110765919A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102916931A (en) * | 2011-08-03 | 2013-02-06 | 北京同步科技有限公司 | Multimedia information distributing system and method |
CN102622580A (en) * | 2012-02-20 | 2012-08-01 | 华焦宝 | Face detection and recognition method and system |
US20170193772A1 (en) * | 2015-12-31 | 2017-07-06 | Cerner Innovation, Inc. | Detecting unauthorized visitors |
CN108875695A (en) * | 2018-07-05 | 2018-11-23 | 深圳勇艺达机器人有限公司 | Visit scene interest stream of people's statistical analysis technique based on machine vision |
CN110298706A (en) * | 2019-07-04 | 2019-10-01 | 山东浪潮人工智能研究院有限公司 | A kind of exhibition room visits interaction and takes pictures evaluation method |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113536038A (en) * | 2021-08-03 | 2021-10-22 | 深圳市一朴创意有限责任公司 | Intelligent exhibition hall distributed control method, device, system, equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106816077B (en) | Interactive sandbox methods of exhibiting based on two dimensional code and augmented reality | |
Miao et al. | Neuromorphic vision datasets for pedestrian detection, action recognition, and fall detection | |
KR100355382B1 (en) | Apparatus and method for generating object label images in video sequence | |
Han et al. | Fast saliency-aware multi-modality image fusion | |
CN106682108A (en) | Video retrieval method based on multi-modal convolutional neural network | |
US10097888B2 (en) | Determining audience engagement | |
CN112215171B (en) | Target detection method, device, equipment and computer readable storage medium | |
CN103283225A (en) | Multi-resolution image display | |
CN101535941A (en) | Method and device for adaptive video presentation | |
CN107341139A (en) | Multimedia processing method and device, electronic equipment and storage medium | |
CN111079864A (en) | Short video classification method and system based on optimized video key frame extraction | |
CN113516666A (en) | Image cropping method and device, computer equipment and storage medium | |
CN116962598B (en) | Monitoring video information fusion method and system | |
CN110765919A (en) | Interview image display system and method based on face detection | |
CN113705510A (en) | Target identification tracking method, device, equipment and storage medium | |
CN110662001B (en) | Video projection display method, device and storage medium | |
CN115605914A (en) | Object detection method, object detection device and object detection system | |
CN112488072A (en) | Method, system and equipment for acquiring face sample set | |
CN106056042B (en) | It generates video data transformation and indicates and analyze the method and system of video data | |
CN106303366B (en) | Video coding method and device based on regional classification coding | |
CN115690496A (en) | Real-time regional intrusion detection method based on YOLOv5 | |
CN110968804A (en) | Museum electronic tag propaganda system based on intelligent Internet of things | |
CN114550079A (en) | Image processing method, device, equipment and storage medium | |
CN112733637A (en) | Panoramic video monitoring method for supermarket | |
CN112784631A (en) | Method for recognizing face emotion based on deep neural network |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200207 |
|
RJ01 | Rejection of invention patent application after publication |