Moving Target Detection And Tracking Technology Of Infrared Rosette Scan Imaging And Its DSP Real-time Realization

Because the active millimeter wave and passive infrared system arc complementary to each other, the millimeter wave and infrared combined system is one of the best technical choices in the future intelligent guidance. It greatly improves the level of intelligentiz.ation and fighting competence of weapons, and is widely applied in end guidance systems.This type of systems possesses the advantages of low cost, good portability, shooting facility, and good tracking performance. In the middle or late of 1970s', with the great development of infrared detection technology and the successful use of miniature infrared optic scanner, the infrared guidance technology has gained great improvement. In 1980s', the research is focused on the multi-waveband, multi-color and multi-mode fusing recognition technology.The research of these systems involves in the branches such as infrared physics, multi-dimension random signal processing, image processing and high speed digital signal processing technology, and so on; especially it involves in the feature description, target detection, recognition and tracking for the video image signal.In the dissertation are discussed the infrared video object recognition features, the infrared rosette scan subimage technology, the fast object recognition and tracking technology and the design-consideration of miniature system, etc. Based on those, system experimenting is made and good results are shown.In the dissertation are analy/ed the basic characteristics of infrared radiation, infrared transmission characteristics, the infrared detector, infrared imaging of high resolution and the scan integration. Then infrared field noises model and target-disturbing model are put forward. In the dissertation are also studied the time-space resolution features of the rosette scan system and discussed the theoretical basis for the subimage target detection, recognition and tracking technologies.Furthermore, the multi-frame accumulation infrared subimage target detection model, image space transformation, segmented region recognition and infrared subimage tracking model are proposed in the dissertation. A random field target detection technology for infrared focal plane array (FPA) imaging is also proposed.In the dissertation are investigated the optimal phascshift controlling technology in multi-frame integration and the multi-frame inertiul imaging technology. These technologies improve the overlay ratio and the resolving power of the observation space, and the target correct detection probability.The last target detection technology discussed in the dissertation based on a single frame greatly improves the fast target discovering ability of the system, andindicates the location of the target as well.The application of the neighbor Markov random field has been analyzed in the dissertation. From the angle, the infrared FPA imaging target detection technology based on the adaptive background estimation is also proposed.These theories and technologies discussed above are successfully applied to the relevant projects national defense. The miniature guidance tracking system with high performance and low cost is realized, and has been put in use.