Aimed At The Mirror Field Of View Image Interpretation System Design

The paper studies and designs a system, it's primarily function is watching and recording the information of the aimer's track and aim operation real time, and by handling this information, computing the error of aim operation. The system include subsystem such as Picture of Track and Aim Ingesting System(PTAIS), Data Recording and Processing System(DRPS), a car, and so on.The PTAIS is used to watch and record aim process real time ,and by relevant design of soft and hard, complete spectrum prismatic imaging of aimer's aim process .The DRPS collect the pictures by aimer's track and aim target real time, and using it's subsystem which is Data Processing and Analyzing System(DPAS),Via extracting the position coordinate data of target center and aimed mirror's reticle center of serial pictrues, calculate the distance error of these two points and draw the error curve . According to the results Judge of the skill of aimer.The processing course is designed by the following operations sequence: Image preprocessing -> Tank region segmentation -> Reticle region segmentation -> Tank's barycenter extraction -> Reticle's center extraction->Error calculation. For single image, preprocessing method is median filter ,region segmentation method is iterative threshold segmentation ,then use morphogical image processing method optimize image . Because tank region and reticle region have almost same gray grade , it's difficult to find a appropriate threshold to obtain each region respectively ,so first a appropriate threshold is choosed to obtain tank region and its barycenter coordinate .then another appropriate threshold is used to obtain the connectivity region of tank and reticle ,after subtract tank region , obtain reticle region and calculate its center coordinate . Using these center points coordinate data we can calculate the error results. Think about the target is dynamic, and pictures are serial, image processing need to be completed as soon as image is received. By recording processing result of pre-image as transcendental information, analyzing the serial images, improving the performance of system.The whole system is tested on several series of simulated image and achieves satisfactory results.