Study On Testing Defects Of Artillery Chamber

The current defects testing technology of artillery chamber is characterized by complex operation, low testing speed and low precision. Based on CCD imaging technique, this dissertation develops a defects testing system of artillery chamber by applying the digital image processing technology and pattern recognition technology. This system deals with the influence of the rifling better, recognizes kinds of defects and calculates their area. It also supplies rusts with the methods of texture analysis and grading. Image processing is the core of the system. It consists of image pre-processing, treatment of rifling, image segmentation, feature pick-up, defects recognition and texture analysis of rusts. Image pre-processing aims to lessen the effect of noise. Analyzing the noise of all kinds of defects and comparing all kinds of processing methods, the system chooses Median Filtering Method. Rifling-treating is a key technique because rough rifling adds difficulty to image processing. So the rifling-treating uses the method of area gray-scale average, which makes up the incontinuity of image. Image segmentation aims to separate the defects from the background. After the analysis of the gray scale of them, the binary method is used. The key point of the method lies in the selection of the threshold. So the selection of the optimizing threshold is an important portion in this dissertation. Defect recognition is the main destination where the research work falls. Besides, the dissertation lays stress on the analysis of the characteristics of defects in artillery chamber, and recognizes the defects making use of its area value and the ratio of short diameter and long diameter. The dissertation gives the standard thresholds after many experiments and compares the results from the standard thresholds with the manual recognition results. It shows that the system can recognize spot, nick, scrape and rust one hundred percent. This dissertation employs gray-scale concurrence matrix as a tool to analyze texture of rusts as well. The laboratorial and industrial experiments show that this testing system enjoys the simplicity of operation, higher speed and higher precision.