In biometric technology, face recognition has incomparable advantages. The video analysis intelligent alarm system based on face recognition can capture the face, human shape and other characteristics of different objects such as fugitives, suspects, controlled personnel and out of control personnel in different monitoring scenes in real time. By comparing with the file characteristics in the template library, the identity landing of the monitored object and real-time early warning can be realized.

1. Rapid development of biometric technology

Biometrics can be roughly divided into physical features and behavioral features.

Physical features include: fingerprint, vein, palm shape, retina, iris, human smell, face shape, even blood vessels, DNA, bones, etc; Behavioral characteristics include signature, voice, walking gait and so on.

At present, biometric technology is becoming more and more mature, and has achieved good application in various fields of economy and society. It has significant advantages in solving security problems and improving work efficiency. In the future, with the integration and diversified application of multiple biometric technologies, it will promote the popularization of products and solutions related to multiple biometric technologies in China, actively stimulate the growth of market demand, and form emerging industries across biotechnology, security, it and other fields.

2. Face recognition is widely used in intelligent analysis

Based on digital and networked video monitoring and image processing, computer vision and artificial intelligence, the intelligent video monitoring system automatically analyzes the video content by using the video real-time intelligent analysis equipment or intelligent analysis software deployed in the monitoring system, finds the abnormal information in the monitoring image, sends out early warning in time, and assists the monitoring personnel to deal with the crisis at the first time. At present, intelligent analysis technologies include behavior analysis, feature recognition, video diagnosis, classification and statistics, etc.

After several years of rapid development, the application scope of video intelligent analysis is mainly concentrated in high-end markets, such as military, government, justice, finance, transportation, supervision, nuclear power plants, airports and ports, and then to middle-end markets such as industrial facilities, education and medical treatment. Now it has penetrated into residential areas, retail stores, warehouse management, logistics and other civil markets. It can be said that the application of video intelligent analysis in China’s intelligent security has bloomed.

3. Difficulties of face recognition embedded video surveillance technology

With the further development of face recognition technology, face recognition technology will further integrate with digital monitoring system, and become another application field of face recognition technology.

With the in-depth development of domestic safe city and smart city projects, the high-definition of urban monitoring has been further popularized, and the number of cameras has increased on a large scale, which greatly reduces the obstacles of face recognition in data acquisition and improves the quality of face recognition. The combination of face recognition and monitoring technology has been reused in the judicial system, which meets the urgent needs of the public security system for urban monitoring, fugitive pursuit, blacklist investigation and other functions, which will help popularize face recognition technology.

Face recognition is divided into three categories: identification, verification and watchlist.

Identification is the process of determining the identity of the unknown person by comparing the face image of the unknown person with the face image of the known identity in the database. It is one to many face recognition; Verification is the process of verifying whether the identity declared by the recognized person is right or wrong by comparing the face image of the current recognized person with the face image declared by the recognized person in the database, which belongs to one-to-one face recognition; Monitoring target person recognition is a face recognition system that detects whether a sample with unknown identity is on the monitoring list, which belongs to one to many face recognition.

In recognition, monitoring target recognition is the most difficult. The main reason is that the non active recognition methods and scene differences, and the changes in posture, illumination, age and face resolution seriously affect the face recognition rate. Many research units have been studying how to apply face recognition technology to monitoring. Under the current conditions, the first consideration is still to capture a good face image.

In addition to the efforts in algorithm, from the perspective of the system application itself, the following problems need to be solved: system structure, face resolution, pose, illumination and age.

Leave a Reply

Your email address will not be published. Required fields are marked *