基于图像拼接技术的农药精准对靶系统应用研究-通信与信息系统专业论文.docxVIP

Subject : Based on Image Stitching Technology Application Research of Pesticide Smart-Targeting System Specialty : Communication and Information System Name : Jin Ruiling (Signature) Instructor : Li Baiping (Signature) ABSTRACT Forest diseases and insect pests is our country ecological environment construction and the afforestation of serious constraints, for the prevention and control of major means for chemical control before, but this method is of very low efficiency.Therefore, it is very necessary by using sensor technology, developed a real-time precision pesticide has on the technology of target, but because of tall trees, tree image acquisition camera and cant get enough distance between the trees, so in the process of pesticide precision of target spray technique, can only be collected images of the visual range, small cant get the whole image of the trees. The core work of this article is in view of the above problem, introduce the image matching technology to the solution of the problem.Based on the research of the classic image splicing technology, this paper explored a set of practical application demand for tree image mosaicing method, eventually be able to quickly and accurately to two local trees image spell a complete synthesis of trees, for pesticide accurately on the target system of a narrow view of the trees image acquisition problems provides solutions, provide the basis for the real-time image mosaicing trees made technology. First of all, in this paper, the image stitching involved various key technologies are analyzed, and explore a set of Harris detection algorithm, the normalized cross-correlation matching algorithm (NCC) and RANSAC algorithm of combining the efficient suitable trees image mosaicing method, finally got a smooth, seamless large trees spliced images. Second, trees on a single set of static image stitching software analysis system is established.Trees in the system, a single set of static image splicing process will be di