| As a branch of computer science, the machine-vision technology developed rapidly in past dozens years. It is a technology of gain image and video data from nature used of digital input device, convert the data into digital information, then it is researched used by computer. The machine-vision technology is the synthesis technology of computer science, optical technology, automation technology, pattern recognition, artificial intelligence and so on. It has great advantage and development prospect that the intelligence recognition based on machine-vision instead human's vision .It has been widely used in the fields of industry, agriculture, military affairs, and national defense and so on.First, this paper talked about the development of the machine-vision technology based on looking up large amount of document, then it was provided that the information of application in the fields of industry, agriculture, health and medical community, national defense military affairs and so on. Last, it was researched that the machine-vision technology applied in field of ammunition test equipment.This paper had introduced a few algorithms in common about digital image processing, an algorithm for ellipse detection had been brought forward based on edge-tracking and Hough transformation, made use of the principle of the line between dual points go through the center point in the graph with two axis, first, it was detected the edge for ellipse graph and made the edge closed, then got the slope for every point on edge, find the dual points with same slope, link the dual points, and accumulate every point in a 2-D accumulator , the maximum point must be the center point, then calculate the length of the semi-major axis and short half axis, and the angle of the major axis, gained the 5 parameters of ellipse. It was simulated used of MATLAB; the rate for ellipse detection based on this algorithm was almost 20 times fast than the traditional Hough transformation..In this paper, it was researched the scheme that the measuring equipment for typical fragment's windward area, the principle of measurement the fragments windward area was simulated by 'regular icosahedron's method', got the conclusion of that it need measure the area in 16 different directors, and calculate these angles, then presented the structure, function and working process of the measuring equipment. The method was studied to improve the calculation's precision and working efficiency. The 4 images of the fragment lying on the revolve table with bright background were gathered at the same time used 4 CCD cameras in different directors. The revolve table rotated by the computer controlled. So the table only rotated four step, 16 images could be gathered, the efficiency could be improved very high, and the measuring time could be reduced. In order to measure the fragment with different dimension, improve the measure precision, the CCD cameras could move back and forward on the ball bearings guideway, controlled by computer, so the image of target could full of the viewing field. The lens could move to adjust focus, controlled by computer too, The cameras and the lens were spurred by screw, move on the ball bearings guideway back and forward driven by DC motor controlled by computer, the distance recorded by grating ruler real time. The method of calculate the fragment's area had been studied, the difficulty of calculate the image's area had been analyzed, the method was studied to improve the calculation's precision and work's efficiency. It was presented the application of image fractionize technology, template technology and VC++DLL (dynamic interlink library) technology in the equipment. The result has indicated the studying achieved the design plan; the equipment was able to measure the typical fragment's windward area quickly and accurately.ICT (Industrial Computerized Tomography) was an equipment that could detect the inner structure of the object used of high energy x-ray orγ-ray. It could get the inner 2-D image of the object quickly, intuitive and accurately. It could detect the inner density and fine structure of the object. But the ICT image always bring the problem of image was blur, border and detail were vague and so on because of various factors such as electron device noise, measure method, ray dispersion and so on. So the ICT image must be enhanced, the vision effect must be improved, then got the objective, accurate conclusion based on the target's characteristics of gradation value and form. In this paper, with the help of the research detection method for ICT image's defect, they were proposed that the image enhancement method based on wavelet transform image fusion and mathematics morphology. The experiment result showed that recur to digital image processing technology, the ICT image's visual effect could be enhanced, be propitious to the object of detect the ICT image's defect.This paper studied the image sequence of movement track which using the remote shooting to get the moving bomb, missiles and other flying targets, analyzed the characteristics of these images, and studied the method of the target automatic identification. Due to include low SNR(Signal-to-Noise Ratio) images of the weak objectives, targets were difficult to identify. Therefore, based on the block-matching technology to proportionate image, it could obtain matching images of unanimous size. Based on the subtracting technology of the cumulative energy to the background, it carried out image sequence average and got background image, from the image sequence subtracting the background image, it could obtain image sequences which contained target and noise. Then the binary images were obtained used of adopting threshold segmentation technology to break up the image sequences which contained target and noise after removed the background. Then the noise were removed and target were detected based on correlation of target between frames. The experimental results showed that this method could effectively detect small target from the sequential images with low SNR.In this paper, it was researched that a new measuring method for velocity of fragments when warhead exploded based on laser screen. The current measuring method for velocity of warhead fragments had been introduced; the problems in the transitional method which contained high-speed photography and contact-type area intercepts method had been analyzed. There were a few problems in the traditional contact-type area intercepts method; because the flying fragments would be obstructed by the screen and the fragments' speed would be affected. So the screens in different distance couldn't place on line, they were staggered each other. It was not the same fragment which hit the different screens. The data each screen collected was the fast fragment's which hit the screen, so it need assume that all fragments fly with same speed. Obviously, the assume object was wrong. In practice, the abnormal data always appeared. Another, the screen used in contact-type area intercepts method only used once, a lot off screens were mangled in one warhead explosion test, and the abnormal data witch from gravel stone and grain sand couldn't be removed. So there were a few problems with inaccurate, inefficient and high cost in the contact-type area intercepts method.For x-ray or laser high-speed photography, the most problem were: (1) the equipment was very complicated, it was very difficult to use on open country. (2)There were a lot of fragment's shadow on one picture, it was very difficult to identify the same fragment in different picture; (3) the picture's effect was poor because of the working field's condition with high temperature, high press and high shock wave, so it was inevitable to bring personal error when a person detected the target, and because the dimension of the viewing field was limited, the puny personal error would be zoomed out in the identifying result; (4) The film could be washed only after the test was over, and the data could be gained after the file was washed, it couldn't get the test data real time.It was proposed that a new measuring method for velocity of warhead fragments based on 6 sky screens, though it was a non-contact measuring method for velocity of fragments, but the equipment was very numerous and jumbled, and the data processing was very complicated, it was unsuitable to use on open country.Therefore, a non-contact measuring method based on laser screen was proposed. This equipment included two laser screens, a multi-channel chronograph and a computer. When the fragments flied through the laser screen, the multi-channel chronograph recorded the time and position of every fragment. According to the data and the parameters of laser screen, the velocity vector and position of every fragment were calculated. This method could discriminate the single fragment's position and velocity vector used image process technology ,then the velocity vector of fragments was measured accurate. It was helped to calculate the accurate initial velocity and attenuation coefficient of warhead fragments. The abnormal date was charge off automatic. The accuracy was anylised, and the contrast test for bullet's velocity between the laser screen and the sky screen. The test result show that accuracy of the laser screen was 0.5%.In this paper, the application of the machine-vision technology in the field of ammunition test equipment was studied; include ammunition ICT image defect detection, measuring equipment for typical fragment's windward area, identify small target from the sequential images with low SNR, a non-contact measuring method based on laser screen. The result showed that in this field, the machine-vision technology could be used widely. It could be applied in many test equipment, achieved the accurate, online, real time, non-contact test process. The autoimmunization of test equipment would be provided; the weapon test equipment development would be expedited.
|
PDF Full Text Request