Research On Technologies Of Infrared Image Target Recognition

As the rapid development of computer technology and infrared imaging technology,image processing based on infrared image gets broad applications, such as militaryguidance, safety monitoring, transportation and medical diagnosis. Target detection andrecognition algorithm is a very important research topic in the field of infrared imageprocessing, it merges many technologies including the science of computer,communication technology, optics, chemistry and so on, besides, it is further importantprerequisite for infrared target tracking.This paper studies on the infrared image target recognition technology, includingimage preprocessing, segmentation and identification of objectives and the design of thesoftware system. For image preprocessing, we discuss the commonly used imageenhancement and de-noising algorithm, propose a filtering algorithm based switchcontrol. In chapter III, we discuss the advantages and disadvantages of three infraredimage non-uniformity correction algorithms. After compared three image segmentationalgorithms, a new multi-feature fusion image segmentation approach is proposed, itsegment targets by combining gray, region information, using mathematical morphologyto get the targets edge more continuous and smooth. According to the characteristics ofthe target in the project, I propose an improved target recognition algorithm based on Huinvariant moments, optimize the training and classification strategy, this algorithm hashigh recognition rate while it meets the real-time requirement very well. We simulate andcompare all the algorithms to verify their feasibility. The article also briefly describes thesystem design and realization of hardware and software of the project.