The Design And Realization Of Remote Automatic Target-scoring System For Unmanned Target Vehicle

This paper discusses the background and significance of remote automatic target-scoring.This paper introduces the classification of automatic target-scoring,compares the advantages and disadvantages of various technologies,and finally adopts image processing technology,and introduces Lab VIEW’s NI-VISION.The main work of this paper is focused on the hardware design of automatic target-scoring system and the software design and implementation based on NI-VISION image processing.The hardware design part mainly realizes the main control,slave control,command communication,image communication,target vehicle drive and target surface design.The main control uses X86 industrial computer and the slave control uses ARM Cortex-M3 as the core processor.The USART serial port and the input and output ports are preset.Instruction communication is based on serial communication between master and slave computers,using 433 MHz wireless module with 1W power,and industrial MODBUS as communication protocol.Image communication is based on LAN wireless bridge,5.8 GHz high-power wireless AP,and WIFI coverage is established within 5 km.The driving part of the target vehicle is designed for the unmanned target vehicle,which is reserved for hardware only,and is not the focus of this paper.The target surface is a square target of 6 meters by 6 meters,and the target surface is equipped with known coordinate points.In the software design part,mainly based on NI-VISION image processing technology,the acquisition of target image,geometric correction,bullet point extraction,target center extraction and position estimation are realized.Image acquisition is based on wireless bridge and surveillance camera.It is transmitted to X86 industrial computer through 5.8GHz channel.Geometric correction validates the designed geometric correction algorithm through the perspective distortion test chart of the predicted coordinates,and brings in the measured picture to maximize the correction of the distorted target surface taken from the lower side angle.Through background subtraction,the binary image of bullet holes is extracted from two adjacent frames with and without bullet holes.Through the measurement of color similarity,a certain threshold is set to get the cross of the target center,and then through the opening operation of the primary morphology,the interference of similar background colors is removed,and the target binary map is obtained.Finally,the NI-VISION-based granularity analysis subroutine is compiled to obtain coordinate data of target center and target bullet hole.Finally,the paper summarizes and prospects the deficiencies and the follow-up research work.