Automatic Weapons System Testing And Analysis Techniques

The dissertation was supported by one pre-research of an important national research project. The systematic study on using computer and digital image processing methods to measure and analyze the automatic weapon's motion state when shooting and the projectile hole coordinates in target paper images after shooting were conducted. The detection of moving objects in a remote-controlled machine gun system was also studied and some corresponding digital image processing methods were proposed, which have been experimentally verified combining with the actual subject and can meet the needs of the project research. The achievements of the paper are as follows:First, the high-speed camera and image processing method were used to solve the problems how to measure the multi-DOF (multi-degree of freedom) motion parameters of the gun body and its automatic mechanism when shooting. A corresponding digital image processing method was proposed. The experiment validates that the proposed method about multi-DOF motion measurement can accurately measure the displacement-time and speed-time state parameters of a multi-DOF moving object.Second, during the accuracy tests of automatic weapons, in order to solve the problem of detecting the distinguishable incomplete overlap projectile holes in target paper images, the watershed segmentation algorithm by using the distance conversion of overlap projectile holes' geometric information was presented. According to the idea of Hough transform, the geometric vertex points were detected for overlap holes in images so as to estimate every hole's center coordinates. The experiment showed that this algorithm can effectively obtain every center coordinates when the number of overlap projectile holes was not large.Third, the motion detection process of the remote-controlled machine gun system was divided into two states, monitor-scan process and target-forward process, according to its characteristics of surveillance and control system. Then two different image processing algorithms for motion detection were proposed. The algorithms were mainly based on inter-frame difference and background difference method. Validated by simulation experiments, the proposed algorithms can meet the real-time requirements.