Face recognition is a kind of biometric recognition technology based on human facial feature information. A camera or camera captures an image or video stream containing faces, and automatically detects and tracks faces in the images, and then performs a series of related techniques on the detected faces, often called face recognition and face recognition.
Face recognition is a popular field of computer technology research. It belongs to biometric recognition technology. It distinguishes organisms from biological features of organisms (generally people). The biometrics studied include biometrics such as face, fingerprint, palm print, iris, retina, voice (voice), body shape, personal habits (such as the intensity and frequency of keyboard strokes, signatures), etc. Face recognition, fingerprint recognition, palmprint recognition, iris recognition, retina recognition, speech recognition (can use speech recognition to identify, voice content can be recognized, only the former belongs to biometric recognition technology), shape recognition, keyboard typing Identification, signature identification, etc.
Three key technologies
First, feature-based face detection technology
Face detection is performed by using colors, outlines, textures, structures, or histogram features.
Second, based on template matching face detection technology
The face template is extracted from the database, and then a certain template matching strategy is adopted to match the captured face image with the extracted picture from the template library, and the face size and position information are determined by the relevance level and the matched template size.
Third, statistics-based face detection technology
By collecting a large number of face positive and negative sample banks for â€œfaceâ€ and â€œnon-faceâ€ images, the system is trained by statistical methods to achieve detection and classification of face and non-face modes.
1, geometric features
From the distance and ratio between facial points as a feature, the recognition speed is fast, the memory requirements are relatively small, and the sensitivity to light is reduced.
2, based on model features
The facial image features are extracted based on the different probabilities of different feature states.
3, based on statistical characteristics
The face image is regarded as a random vector, and statistical methods are used to distinguish different face feature patterns. Typical face features, independent component analysis, and singular value decomposition are compared.
4, based on the characteristics of neural networks
A large number of neural units are used to associate and store facial image features, and accurate recognition of human face images is achieved based on the probability of different neural unit states.
Top ten difficulties
1, lighting problems
Illumination change is the most critical factor affecting the performance of face recognition. The degree of resolution of this issue is related to the success or failure of the process of face recognition. Due to the 3D structure of the human face, the shadow cast by the light will strengthen or weaken the original facial features. Especially at night, the shadow of the face caused by insufficient light causes a sharp drop in the recognition rate, making it difficult for the system to meet practical requirements.
At the same time, the theory and experiment also proved that the difference of the same individual due to different illumination is greater than the difference between different individuals under the same illumination. Illumination is an old problem in machine vision and it is particularly evident in face recognition. Solutions to solve the lighting problem include three-dimensional image face recognition and thermal imaging face recognition. However, these two technologies are still far from mature, and the recognition effect is not satisfactory.
2, posture problems
Face recognition is mainly based on human facial features. How to recognize face changes caused by posture has become one of the difficulties of this technology. The pose problem involves facial changes caused by the rotation of the head around the three axes in a three-dimensional vertical coordinate system, where deep rotations in two directions perpendicular to the image plane cause partial deletion of facial information. Making pose problems become a technical problem for face recognition.
There are relatively few researches on attitude. Most current face recognition algorithms mainly include frontal and frontal face images. When the pitch or left and right sides are relatively severe, the recognition rate of the face recognition algorithm will also be A sharp decline.
3, facial problems
Facial facial expression changes such as cries, laughs, and anger also show the accuracy of facial recognition. The existing technology has handled these aspects quite well, either on the open mouth or on some exaggerated facial expressions. The computer can correct it through three-dimensional modeling and pose correction methods.
4, occlusion problem
The occlusion problem is a very serious issue for face image acquisition in non-match situations. Especially in the monitoring environment, often the monitored objects are wearing glasses, hats and other ornaments, making the collected face images may be incomplete, thus affecting the subsequent feature extraction and recognition, and even result in face detection algorithms. Failure.
5, age change
As one's age changes, a person changes from a teenager to a young person and becomes an old person. His appearance may undergo a relatively large change, leading to a decrease in the recognition rate. For different age groups, the recognition rate of the face recognition algorithm is also different. The most direct example of this problem is the identification of ID photos. The validity period of ID cards in China is generally 20 years. During these 20 years, each person's appearance will inevitably undergo considerable changes. All of them have a great deal of recognition. problem.
6. Face similarity
There is not much difference between different individuals. The structure of all faces is similar. Even the structure of the human face is very similar. Such a feature is advantageous for the positioning of human faces, but it is disadvantageous for distinguishing human individuals using human faces.
7, dynamic identification
In the case of non-combined face recognition, facial image blurring caused by motion or incorrect focus of the camera can seriously affect the success rate of facial recognition. This difficulty has become apparent in the use of security and surveillance identification in subways, highway bayonet, station bayonet, supermarket counters, border inspections, etc.
8, face security
The main method of deception to forge face images is to establish a three-dimensional model, or to embed some expressions. With the introduction of intelligent anti-counterfeiting technology, 3D facial recognition technology, camera and other intelligent computing vision technologies, the success rate of forging facial images for recognition will be greatly reduced.
9, image quality problems
Face images may come from a variety of sources. Due to the different acquisition devices, the quality of face images obtained is also not the same, especially for those face images with low resolution, high noise, and poor quality (such as those shot by mobile phone cameras. Face images, remote monitoring and shooting pictures, etc.) How to perform effective face recognition is a matter of concern. Similarly, the influence of high-resolution images on face recognition algorithms needs further research. Nowadays, when we use face recognition, we generally use the same size and sharpness of face images. Therefore, image quality problems can be basically solved. However, in the face of more complex problems in the real world, we need to continue to optimize the process.
10, lack of sample
The face recognition algorithm based on statistical learning is currently the mainstream algorithm in the face recognition field, but statistical learning methods require a lot of training. Because the distribution of face image in the high-dimensional space is an irregular manifold distribution, the available sample is only a small part of the face image space. How to solve the statistical learning problem under small sample needs to be further studied. Research. In addition, the face image libraries currently involved in training are basically images of foreigners. There are very few face images libraries for Chinese and Asian people, which makes training facial recognition models more difficult.
Face recognition application dimension
Two dimensions of dynamic scene
The definition of 1:1 is a role of judgment. The application scenario is actually financial and personal identification. The characteristics are more precise and safe. So now everyone is either Alipay or the bankâ€™s identity comparison and real name business. Will use 1:1 face recognition.
1:N is more likely to be found in a database or in a base library. It is a process of identification, a dynamic, or a non-cooperative scenario. Saying that in the course of security, I went to capture fugitives. I went to catch fugitives who could not allow fugitives to see the camera. It is also impossible for our VIP customers, employees, and members to perform operations on the camera head in commercial scenarios, so it is a dynamic and non-cooperative scenario.
Four dimensions of business scenarios
First, the tray is large enough to support the long-term development of the company.
Second, data flow back.
Third, whether it is the use of high-frequency scenes and high frequencies.
Fourth, whether it can be copied or not, whether it can be changed from 1+0 to 1+N can improve efficiency.
Three dimensions of the visualization system
First, personnel traffic management.
Second, the convergence of sensor networks.
Third, the integrated part of commercial real estate + new retail.
Face recognition application
Applications: finance, justice, security, border inspection, aerospace, electric power, education, medical care, etc.
Four potentials for commercialization: gates, traffic, banks, mobile phones
Face recognition categories and their corresponding application areas at a glance
Application scenario overview
The financial sector:
1, face recognition autonomous terminal
Manual review, self-opening cards, business changes, password resets, and other personal services.
2, mobile finance, sales
Verification of remote identity verification, in two aspects: verification of user identity and portable devices with face recognition systems required for financial institutions to conduct business at home.
3, counter system
Face-to-face verification, used in banking, insurance, securities, and other financial institutions to open accounts, and other services.
Three key points for domestic airport applications: first attempt, boarding, full intelligence
1. The first attempt of Beijing Capital Airport in 2009 was the first step for domestic airports to begin to recognize face recognition technology. However, due to the level of face recognition technology at that time, magnetic cards had to be used for cross-validation to ensure uniqueness of identity. In terms of recognition speed and accuracy, face recognition technology at the time and face recognition technology after deep learning intervention were not at one level.
2. In 2014, Nanjing Lukou Airport attempted to apply face recognition technology to boarding for the first time. Although it was also limited by the level of technical business at the time, it was unable to achieve self-service clearance, but it provided ideas for subsequent applications. And experience.
3. In December 2016, the comprehensive intelligence of Yinchuan Airport marked a new stage in the level of airport intelligence. In addition to security check-in and self-check-in, face recognition and related computer vision technologies are applied to dynamic deployment, flow guidance, smart aviation display, VIP welcome, track retrieval, cleaning alert, etc., for face recognition in 2017 Technology laid a good foundation for the outbreak of airport applications.
China Southern Airlines - China's first airline company to use face recognition technology! The CZ3384 became the first flight to use new technology for boarding. Travelers do not need to hold the boarding pass, and can quickly pass through the boarding gate by brushing their faces.
Chinese style crossing the road:
1 Use face recognition to solve cost problems.
2 Adhere to the administration by law to prevent extraterritorial punishment.
3 Solve road rights conflicts and avoid sports enforcement.
According to the Jinan police, the face recognition system is mainly used to capture pedestrians and non-motorized drivers who drive red lights, and can also clearly image at night. Pedestrians were â€œgrabbed by the currentâ€, and the short video of the red light and the enlarged head were directly exposed on the display of the intersection and presented to the public.
In addition, this set of equipment is also connected to the resident information system. Personal information such as the name of the offender and the identity card number identified through face recognition will also be displayed on the electronic screen.
After the implementation of the face recognition system in Jinan, a total of more than 6,200 illegal pedestrian and non-motor vehicle violations were captured in one month. With the deterrence of "black technology," the behavior of the red light has been effectively curbed, and there has been an intersection where the number of red lights per day dropped from over one hundred times to a dozen times.
In Chongqing Jiangbei, since the face recognition system was put into trial operation, the law-abiding pedestrian crossing rate has increased from 60% to over 97%.
Hidden Trouble: The disclosure of personal information to the public involves the disclosure of personal privacy. Experts suggest that for information collection activities such as face recognition, an announcement must be made to the public in advance to inform the public that it has entered the public information collection area and illegal activities will be photographed and exposed. This not only satisfies the publicâ€™s right to information, but also serves as an alert. Role; The collection of information, appropriate technical processing, privacy should not be open, it should be covered or not open.
Fundamental: According to interviewed experts, the difficulty in crossing the road due to the unreasonable setting of transportation facilities is often the main reason leading pedestrians to run into the red light. Some urban road network planning is irrational, and the construction of main roads is emphasized. The density of branch roads and secondary roads does not meet the requirements, causing both pedestrians and non-motor vehicles to be collected on trunk roads; the intersection of traffic lights at some intersections is unreasonably timed. To cross the road according to the rules, you must have enough patience and fast enough speed. Only by comprehensively managing and resolving conflicts between human and vehicle "road rights" can we fundamentally solve the "Chinese style crossing the road."
Education: Candidates for identification and identification, campuses, and dormitory access management.
Example: In the 2016 college entrance examination, there have been many provinces such as Beijing, Sichuan, Hubei, Guangdong, Liaoning, and Inner Mongolia that have adopted the â€œface recognition + fingerprint recognitionâ€ biometric identification technology to confirm the candidate's identity and prevent quitting and cheating. With the expansion of pilot regions and various fields and the maturity of operating models, the Industry Research Institute expects face recognition in 2017 to usher in large-scale popularity.
(1) Face capture and tracking.
(2) Face recognition calculation.
(3) Face modeling and retrieval.
* Field Face Recognition Products
Use is mainly reflected in two aspects: on the one hand, the use of dynamic face recognition systems in the background, on the other hand, the use of front-end face recognition handheld devices and human comparison cards.
1. Community medical examination application
When the community uses digital physical examination equipment (electronic blood pressure monitors, body scales, blood glucose meters, etc.) to transmit data to the digital medical records or health records, plus the living and facial information of the patients, they are stored in real time. Identification. Each record after the completion of the unique identity will be recorded, so that the status of the patient can be quickly fed back to the doctor and the patient, and the best treatment plan can be obtained.
2. Application of medical institutions at grade 2 and above
By setting up a face recognition system in different scenes such as a kiosk, a window, and a doctor's office, the identified face information is used as an information search entry, and the patient's information file is associated with each other, and then the face can be retrieved and the clinic record can be retrieved.
Smart City Area:
1. Pension management
The use of face recognition technology can effectively conduct personnel verification and reduce the loss of pensions.
2. Tax Certification System
Through the face recognition technology, the system automatically compares the lens intake image with the portraits in the * department identity information, realizing real-name authentication in real time. It not only effectively relieved the pressure of the window tax personnel, improved the efficiency of taxation, but also enhanced the real-name taxation experience and reduced the tax-related risks.
4. Community Management System
In the smart city, taking the smallest unit community in the city as an example, non-combined face recognition can help the property management department to provide owners with more friendliness in terms of visitor management, property notices (utility notices, garage information, etc.). Natural life experience.
5. Building Access Control System
The face recognition intelligent access control system can improve the security of buildings and families by constructing an identification system with intelligent management functions and combining advanced face recognition algorithms to accurately and quickly identify faces and open access controls.
6. Candidate Identity Management System
Based on the special needs of the examination industry, an application system project integrating computer, communication, network, face recognition technology, database, and other diversified technologies will provide examination organizations with functions such as information extraction, identity verification, and management of candidates' ID cards. For an efficient and fair examination environment.
7. Drivers' identity information authentication and safe driving management system
Including on-site verification, student identification, getting on and off the car, driving time control.
8, smart diet management system
The system performs face recognition when the students are having dinner, records the foods eaten by the students each day, and then compares and analyzes the results of the hospital's physical examination to obtain dietary adjustment opinions. The students are required to record foods that have been wasted for more than a specified amount, and then continue to optimize the dishes. To achieve the goal of adjusting and optimizing the student's diet.
9. Business Intelligence Analysis System
The face recognition system can make full use of machine vision's feature recognition and induction capabilities of the human face, and take the customer's gender, age, and mood as the corresponding characteristics of business needs, and push targeted and real-time customer-interested content to target businesses. The customer group diverts and accurate sales; On the other hand, through the observation and learning of the interest content of different groups of people, the matching accuracy of the push content of the target group is gradually improved.
We are good reputation China Paper Making Machinery manufacturer, we can mae different kinds of paper making machinery ( Kraft Paper Machine , Corrugated Paper Machine, flute Paper Machine , A4 Paper Making Machine etc.) as well as the pulping system and auxiliaries. Rewinder machine is one of our unique products, this series of underfeed Paper Rewinder Machine is shaftless-type, double supporting rollers. The whole structure is split type. Drive control mode: DC motor drive and AC motor drive. Tension control mode: electrical direct tension control and electrical indirect tension control.