Extraction Algorithm Of High-speed Weak Target Basing On Linear CCD

Target position measurements in the shooting range test is a very critical parameter, the current test methods use target screens test system, the line array CCD intersection test system,etc, these systems need to target identification and there is a deviation in the actual extraction because of the particularity of the dynamic parameters in the recognition process, a blank template based on deconvolution is proposed for this alternative method and extraction algorithm of high-speed weak target basing on linear CCD is studied.Based on double CCD intersection space disposal relations, double CCD intersection measuring system model is established, the analysis of lens focal length, the baseline length, the height of the ballistic, the distance between CCD pixels, the angle of main axis and horizontal direction, show how to calculate in wire array CCD camera imaging surface imaging point coordinates formula, and can be projected to the exact location of the target in the space. Using digital image processing method of filtering, subtraction, division of the high-speed flying image processing and analyzing of the disadvantages of various methods; cross-correlation image method and blank template substitution method based on the deconvolution are proposed, which research projectile row and column template respectively operate correlation and deconvolution of high speed projectile image rows and columns, then extract the minimum space row and column coordinates information, by comparing with the surrounding background gray identify real projectile target imaging position and recovering, the MATLAB simulation can validate the algorithm. Compared with commonly used image processing methods, cross-correlation image method and blank template substitution method based on the deconvolution can improve the accuracy of the projectile object extraction and avoid confusion of the real target and noise point; according to CCD imaging characteristics and high speed projectile target and the background gray feature of image, blank template substitution method of deconvolution study the correlation between target and background pixels, which made target and the background get significantly separated and the target characteristics get more evident.By the study of blank template substitution method of deconvolution of high speed small target extraction algorithm for all types of flying projectile image processing provides a more simple and convenient method, which has been applied in the actual specific test conditions and laid a good foundation for subsequent calculation of various parameters.