Study And Implementation Of Auto Focus And White Balance Algorithm In Camera

In recent years,digital image processing chips with high level of integration and powerful performance are increasing rapidly,so the IP camera appears with the high-speed development of network.Developed from the traditional camera,IP camera is a new product of the traditional camera combined with network video technology.As a core module of the IP camera,image signal processor has a great influence on the camera's performance.Auto focus and auto white balance algorithms are very crucial to the image signal processor,which have a great effect on image quality.In this paper,we focus on image signal processing algorithm of high-definition IP camera based on Hi3516 series' audio and video processing chip.Then auto focus and auto white balance algorithms are researched in detail.An auto focus algorithm is designed and improved on the platform of Hi3516 C,and an auto white balance algorithm is developed on the platform of Hi3516 A.Auto focus keeps making a real-time definition evaluation of images captured by optical lens and sensor,then gives a feedback to the camera to control lens move,keeping adjusting until the image definition reaches optimum.A new auto focus method with higher precision,higher speed and strong stabilization is designed for Hi3516 C high-definition IP video conference camera in this paper.Firstly,the specific algorithms of each module in the auto focus system are introduced.Then,solutions for the main problems during the implement process to improve the performance of auto focus.Finally,the algorithm is tested in conference scene.The experimental results indicate that a high video quality of subjective perception and strong stabilization can be achieved by the proposed method,and the average focusing time of the proposed method is 0.90 s,specifically,the average focusing time in small zoom,middle zoom and large zoom are 0.60 s,0.85 s and 1.25 s respectively.It can satisfy the high-definition IP video conference camera requirements and has been applied to terminal products.Auto white balance is an important technique in camera,which is used to remove the influence of different light sources from images to make the images captured by camera close to the subjective feeling of human visual.The auto white balance module of the Hi3516 A chip has a good balance result in the ordinary scene,but appears color correction error in the monochromatic scene.Aiming at the defect of auto white balance algorithm ofthe Hi3516 A high-definition IP surveillance camera,a new auto white balance algorithm based on color temperature curve is designed in this paper,along with the auto white balance algorithm of Hi3516 A to guarantee the whole performance of high-definition IP camera.The experimental results indicate that the proposed algorithm has better white balance performance in monochromatic scene with good adjustment stabilization and fast convergence,and can effectively work together with the auto white balance algorithm of Hi3516 A.It successfully solves the white balance problem of Hi3516 A,as a result,the whole white balance performance of the high-definition IP surveillance camera is got improved.