High Definition Auto Focus And Its Application Research In The Network Integrated Camera

In recent years,with the development of digital and network,network camera has become an important component of the security monitoring system.High definition network camera(HD_IPC)has the characteristics of high definition,high interactivity and intelligence compared with the traditional analog camera.Therefore,the high-definition network camera is widely used in a variety of monitoring occasions,such as banking,security,highways,etc.,and has become an important security industry terminal product.In the network camera,the integrated network camera occupies a very important part,which can meet the needs of high monitoring in different scenarios.Auto focusing technology is one of the core technologies in the integrated network camera,and with the development of digital and chip technology,it has entered a new stage.This paper focuses on the research of high definition auto focus and its application in the integrated network camera.Auto focus algorithm mainly includes three core algorithms:image definition evaluation function,peak search algorithm and zoom tracking algorithm.This paper proposes the qualitative evaluation criteria and the quantitative evaluation criteria of the image clarity evaluation function.Then,the existing and commonly used definition evaluation functions are summarized and classified.Based on previous common evaluation function,this paper details the image clarity evaluation function based on wavelet analysis,and carries on the improvement and optimization,and present the image clarity evaluation function based on the fusion of wavelet transform and gradient function.The algorithm proposed in this paper has better sensitivity and anti-jamming performance than other evaluation functions,and has better comprehensive performance.According to the ideal characteristics of the focus curve,this paper improves the blind hill climbing algorithm and puts forward the variable step peak search algorithm based on new judgments.This method divides the different parts of focus curve into four states:Initial,Coarse,Mid and Fine.Different states have different step motor precision,it can significantly reduce the running time of the algorithm.This paper researchs the existing zoom tracking algorithm,and proposes the IFZT algorithm based on tht improvement of FZT algorithm.Compared with the FZT algorithm,IFZT modifies the correction criterion of the feedback correction point,and makes a special processing for the large defocus state,and removes the relatively complex PID algorithm.At last,this paper establishes IPC experimental platform based on Hisilicon Hi3516A chip,writes the back-end service program,and carried out the test experiment.The experimental results show that the experimental platform can realize the automatic focusing function accurately and stably,and meet the design requirements.