Study On Extraction Technology Of Low Contrast Extended Target In Sky Background

With the development of military, aerospace and other fields, the target detection and recognition technology have attracted more and more scholars' attention, and they also carry out the in-depth research on it. Because of the influence of various factors such as weather, equipment and so on, resulting in the acquired targets appear to be local or global fuzzy, at the same time, it leads a difficulty to extract the effective feature information, so it is not conducive to have accurate in-depth research for the target.The research content of this paper is mainly about low contrast targets detection, the technical route used in this paper is about the key issues on the study of image enhancement and edge extraction. This paper has a further study on the problem, the main research results are:1 A corresponding target curve about the range of image contrast and the target complete extraction is explored. Through the analysis and calculation about local image contrast and edge extraction integrity, the performance curve of edge detection algorithm to extract the extended targets has been proposed, which not only quantized defined the concept of low contrast extended target, also can automatically determine whether need to do some preprocessing on the target according to the performance curve, such as image enhancement, in order to improve the quality of target extraction.2 A local image enhancement algorithm based on pixel spatial location is proposed. In view of the low quality targets in low contrast situations, this paper analyzes several different enhancement algorithms, and proposes an improved algorithm. It uses pixel spatial location to determine the scope of the target area, and then enhance the local area contrast of image according to the scope of target, getting an image which has a good visual effect after contrast enhancement to the images. At last, this paper verifies the effectiveness of the proposed method by comparing with the objective evaluation results of several other enhancement algorithms.3 A method of automatically determine the continuity and integrity of the edge extraction is proposed. For the problem of edge continuity, the solution is to extract the edge information after filling the target edge first, then determine the continuity by using edge degree of continuous and area ratio which filled before and after: if it satisfies the two conditions at the same time, the edge is continuous, or the next step to make edge connecting will be done directly, in order to achieve the complete closure of the edge information finally. The method solves the way that human intervention to determine the continuous of edge or not, improving the intelligent degree of low contrast targets extraction.After a series of the image processing and a large number of tests, it is shown that the approach used in this paper can effectively extract the edge of low contrast targets more complete, and lay a solid foundation for the follow-up of high-precision positioning and tracking.