Research On Dynamic Scene Image Measurement Based On Microarray Lens

The technology of measurement has important applications in many areas,including urban traffic control,intelligent monitoring in industrial production,and safety monitoring.Several different types of distance measurement devices have been used in the past,they can be categorized as active and passive.Active measurement mainly include ultrasonic distance measurement,laser ranging and infrared range finder,etc,active devices send signals into the environment and receive them back,information related to distances is retrieved by analyzing and comparing the outgoing and returned signals.In fact,it is relatively easy to implement,but the results are easily affected by environmental factors,and the requirements for equipment and environment are higher.However,the passive measurement technology based on image and video can largely exclude the instability of environment and equipment,and it can get the three-dimensional information of target accurately,which has wider application prospects.The research is based on the images taken by a 3x3 microarray camera to identify and measure the target in the image.Compared with the traditional camera,the array camera can get multiple images with different views at one shot,which has great advantages in stereo matching and 3D reconstruction.The main work of this paper includes camera calibration,stereo matching,3D reconstruction,target detection and tracking,and three-dimensional information measurement.The main work lists as follows:1.The basic theory of computer vision is introduced,including camera imaging principle,the camera calibration method proposed by Zhang,commonly used stereo matching method,parallax ranging principle and target detection and tracking in dynamic scenes.2.The plane calibration method proposed by Zhang is used to calibrate the microarray camera,and the internal and external parameters of 9 lenses are obtained.3.The depth information is obtained through the NCC stereo matching method,and the operation speed is improved by differential summation theorem.Then,the target scene is reconstructed by 3D reconstruction to get the complete scene information.4.The gray level projection method is used to stabilize the dynamic scene,and the background difference algorithm based on Gauss mixture model is used to detect and track the moving objects appearing in the image,and calculate the three-dimensional information of the target including length,width and height in real time.Based on the method proposed in this paper,experiments are carried out on the images acquired by the microarray camera,which can effectively achieve measurement result of human height in real time,and the measurement results have high accuracy.Experiments show that the microarray camera has a good effect in image measurement.The microarray camera has obvious advantages compared with the traditional camera,and has broad application prospects.