Research On Image Replacement Based On Microarray Camera In Dynamic Scene

Image replacement technique in dynamic scene is we replace one or more objects in each frame of a video,according to user needs with other interested things of the image or put it in a new background to get a new set of image sequences,and then we can get a new video.In recent years,the related technology of computer vision and digital image processing has developed rapidly.Among them,image replacement technology is widely used in various fields such as film and television special effect production,graphic design,leisure entertainment,game scene making and so on.At the same time,it can greatly reduce the cost of production while meeting the user's personalized demand for the image.The microarray cameras can capture static images and video clips that are attached to the information of the depth of field,which can enlarge the field of view and obtain more image information,moreover,it has small volume,can reduce the occupancy of space,and has great application prospects in some small mobile devices,such as mobile phones,flat panels,etc.But at present,there are relatively few researches on array cameras in China.In this paper,a method based on microarray camera using depth information of the image to replace the image in dynamic scene is proposed.The main work is as follows:1.This paper uses the Zhang Zhengyou camera calibration method to calibrate the 9 sub lenses in the microarray lens module one by one,and obtains the related parameters of the microarray camera.2.The basic principle and common algorithms of moving target detection are introduced,and the position of the replacement target can be determined by detecting the moving objects in the video.3.The principle of stereo matching based on graph cut algorithm is elaborated.According to the constraints of the gray consistency of the 9 images and the depth smoothness of adjacent pixels,the corresponding energy function is constructed,and the depth map of the target image is restored by using graph cuts algorithm.4.Based on the image replacement method,the video is divided into frame processing to obtain the depth map of each frame,and the replacement work of the video image is realized.The research of image replacement in dynamic scene based on microarray camera is still a new research field.The algorithm has the advantages of accurately extracting the replacement target and effectively dealing with the front and rear occlusion problem of objects in the video frame,reducing the tedious interaction operation and reliant on operator experience.It is also expected to be applied to the photography and image processing technology of mobile camera.