Image Guidance Technology And Experimentation Research

Precision strike weapons in Modern war has been articulated more and more demands, precisionguided technology is the core technology of the precision strike weapons, and the automaticidentification and tracking technology based on the image is a important problem in badly need ofsolution, and also is the research focus in the field of computer vision. This paper is a very importantpart of the research project called "Heaven and Earth Battlefield Network Simulation Platform", itstudies automatic identification and tracking technology, calculation and test systematically. The mainresearch work includes:First, the image preprocessing. In addition to the commonly used denoising algorithm, make theimage enhancement in bad weather conditions as the important content of the guidance for the firsttime, and combine detection algorithm based on significant and the new adaptive filter algorithm toform a new optimization algorithm in image cutout..Then researched the target recognition and tracking algorithm based on image. Make automaticidentification method based on the Haar feature expands to the battlefield military targets automaticidentification and create classifier training methods based on the AdaBoost algorithm; Combine theKalman filter forecast method and the Camshift algorithms to improve the target tracking algorithm’sspeed and anti-interference ability.Finally, Test the theory above-mentioned. Built automatic identification and targets trackingsystem including software and hardware based on the automatic identification and tracking method inthis paper. Make ARDrone four rotor aircraft as the control target, execute the automatic identificationand tracking target test through a wireless network communication channel.Practical Examples Indicate: the automatic recognition and tracking system builded in this paper,can recognize and track target automatically and has the real-time and anti-interference ability.