Application And Research Of Image Recognition Technology In Intelligent Target

With the development of computer technology in full swing,military shooting training targets have gradually evolved from manual reporting to intelligent automatic target reporting.At present,some military training target reporting methods are relatively backward,and live reporting uses more manual reporting.Targets,troops and soldiers use visual inspection or telescopes to report targets to determine the number of firing rings.Only a few use automatic target reporting and use expensive integrated circuit chips.The target reporting is not accurate enough and has not been promoted and popularized.Therefore,researching an intelligent target reporting based on image recognition is of great significance for improving the accuracy of target reporting.(1)Use raspberry party chest ring target target image to collect.(2)Image preprocessing and feature extraction are performed on the captured shooting pictures.During the experiment,the shape features of the target image were geometrically corrected,and the image was pre-processed using spatial coordinate transformation and gray-scale interpolation.At the same time,image filtering was used to initially eliminate noise,image subtraction,and preliminary bullet hole segmentation.The image difference and target image are binarized to complete the bullet hole segmentation.(3)To determine the ring value of the target image,this system uses Susan algorithm for edge detection and Hough transform application to detect the analytical curve of the image space,determine the coordinates of the bullet point,find the ring value,and report the ring value through the Raspberry Pi.target.Finally,in this thesis,the target is reported by experiments to verify the target.This system can accurately detect the target ring number,achieve target target reporting,and provide help for shooters to improve their shooting level.It eliminates the hidden dangers of manual target reporting,improves the accuracy of automatic target reporting,has better real-time performance,reduces the cost of automatic target reporting,and facilitates the promotion of intelligent target reporting.