| Stations,airports and other transportation sites are the main channels for people to enter and exit the cities.Effectively control the safety of these links,prevent and monitor the suspicious personnel,have significant significance for urban public safety and case detection.At present,these transportation sites basically use manual verification methods for security verification,both accuracy and efficiency are disappointing.With the development of technology,manual verification mode gradually turns into automatic verification mode.This requires the design of identity verification pedestrian gate based on face detection.The identity verification and face detection are automatically performed by networking,which saves time of passengers and improves verification efficiency.Firstly,the overall solution of personnel verification system is designed.The overall system solution was designed.The bottom layer of the system is responsible for information collection,that is,image and data information are collected by cameras,radio frequency identification,etc.the core layer of the system is responsible for face recognition of the collected images;the upper layer of the system is the application layer,which mainly includes logic control of the gate,data transmission and Interface expansion and other functions.Secondly,the implementation scheme of the system hardware level is designed,including information collection,face recognition,system management,external interface,gate control five parts.Combined with the results of the system trial run,the system was optimized to reduce the system delay.Furthermore,the implementation of the system software level is designed.The key technologies involved in the system software level are deeply analyzed and studied,including face image acquisition,face detection and location,image preprocessing,feature extraction,face comparison and recognition.Finally,under the condition of special environment around the gate of the gate,the face detection algorithm based on Adaboost face detection algorithm,Double skin color model under complex illumination and the face recognition algorithm based on hidden Markov model are studied.Aiming at the technical principle of face recognition and face matching,the face recognition algorithm and face detection algorithm are optimized,and the face recognition algorithm and face detection algorithm are selected according to the surrounding environment of the gate channel.The gate channel port identity verification system was tested,and the test operation results were tested and analyzed.The function and performance of the identity verification system have met the expected requirements. |
