The Analysis Of Acquirement Factor And Target Detection Technology Research On CCD Vertical Target

Target detection technology as a key technology of CCD vertical target, it is a prerequisite to measure CCD measuring system can stand as an effective means of testing. The paper describes the status of the research at home and abroad CCD vertical target on the basis of a comprehensive analysis of the effects of CCD vertical target capture factors, focusing on target detection techniques were studied, the main contents including:(1) Based on the linear CCD vertical target measurement system theoretical studies, from the lens focal length, length, angle of the optical axis of the baseline and a horizontal direction, etc. are given through the target location coordinates of the target formula, which is derived by the exact location of a target in space.(2) In this paper, external factors affecting the CCD vertical target measurement system captures quantitatively analyzed from the target characteristic, camera features, environmental conditions, background characteristics, lens performance in five areas.(3) Based on theoretical analysis, this paper focuses on target detection techniques were studied, firstly, according to the CCD camera imaging characteristics, given the similarity calculation algorithm of the target and background. Target energy is very weak in this paper, and the similarity of objects and the background is very high, so the traditional poor shadow method, edge detection and correlation method to detect the presence of defects in the goals,for this defect, SIFT algorithm is proposed based on the energy analysis and the Euclidean distance based on discrete wavelet transform. Euclidean distance analysis based on the energy of background modeling method firstly, and then by analyzing the distance between the template and the image of each frame can (similar to potential energy), while for the presence of false targets, can be obtained by similarity match final the target area; the image algorithm based on discrete wavelet transform SIFT secondary wavelet decomposition, later followed by extraction target by SIFT algorithm feature points, and then combined with the expansion method, projection method and Gaussian blur algorithms, and ultimately can get a more complete destination information. Finally, through the plurality of images containing small targets, including the processing of the robustness of the algorithm was tested, the results compared to the SIFT algorithm based on discrete wavelet transform, energy analysis based on Euclidean distance is more robust high.In this paper, different algorithms for the destination number and the amount of information on how much the merits of the algorithm is measured, to intuitive display of the treatment effects of the different algorithms, the interface algorithm is written based on MATLAB GUI.Through the CCD vertical target of object detection technology research and experiment, which has the great significance to improve the range of the test level.