Font Size: a A A

Automatic Target-Scoring System Based On Image Processing

Posted on:2013-06-02Degree:MasterType:Thesis
Country:ChinaCandidate:X L XieFull Text:PDF
GTID:2246330362974807Subject:Mechanical and electrical engineering
Abstract/Summary:PDF Full Text Request
The current target-scoring in troops live-fire drill is made by manpower, and thetarget-scoring efficiency is low, also hide safety problems. In order to overcome thedeficiency of manpower target-scoring and meet the requirements of modern militarytraining, improve the troops shooting level, the design and implementation of highaccuracy, strong practicability of low-cost automatic target-scoring system hasimportant significance. In this paper, target bitmap is taking as the research object,digital image processing technology is taking as the theoretical foundation, and a suitautomatic target-scoring system based on image processing will be developed by usingmore accurate algorithm of target image processing.To achieve these goals, this paper studies the following aspects.This paper has studied the developing trend of automatic target-scoring system athome and abroad, and then determines the research program and content of automatictarget-scoring system.This paper has studied the algorithm of target bitmap geometric rectification, Inorder to rectify the distortion of target bitmap, quadratic polynomial is used in spacecoordinate transformation of the target bitmap, cubic interpolation is used in gray-scaleinterpolation of transformed target bitmap.This paper has studied the algorithm of extract the information of chest bitmap,and a method to get the center coordinate of chest bitmap and the radius of target ringsis proposed. Eight neighborhoods of laplacian operator using a negative peak, the regionof interests, and fitting circle with least squares method is used in this method. Thispaper also uses a method to extract the effective region of chest bitmap, it is based onOTSU, mathematical morphology and seed filling algorithm.This paper has researched the algorithm of bullet-spot’s recognition and scoring.First it uses a method to distinguish all bullet-spots, and get the information ofbullet-spots, this method is based on maximum entropy automatic threshold method, theopen operation of mathematical morphology, and the improved algorithm of eightconnected domain labeling based on equivalence. Then a method based on theminimum area bounding rectangle of connected component to identify the overlappingof bullet-spots is proposed. At last, this paper proposes a scoring method when the targetbitmap have only a bullet-spot and have multiple bullet-spots. This paper has researched and implemented the automatic target-scoring systembased on image processing. First analyzing the system requirements, making the overalldesign scheme of system, and then configuring the hardware equipment, choosing theNI smart camera as image acquisition hardware, finally developing full-featured systemsoftware.In order to test the automatic target-scoring system developed in this article, it isused to scoring in live-fire drill, and then analyzes the information of chest bitmap andresults of target-scoring which are obtained by the test. The results show that the systemhas high target-scoring accuracy rate, basically meets the target-scoring requirement oftroops live-fire drill.
Keywords/Search Tags:Image Processing, Automatic Target-Scoring, Geometric Rectification, Circle Fitting, Threshold Segmentation
PDF Full Text Request
Related items