| To enhance the photo-electric detection ability, especially the detection ability of infrared dim targets in complex background, of Photo-electric theodolite that is a major observe-control equipment of military range, has very important effect and meaning in enhancing the observe-control ability of equipment of military range. Under this requirement, the infrared dim targets detection technique of Photo-electric theodolite is researched in this paper.Firstly, the background and meaning of research on the detection technique of infrared dim targets in complex background in the Photo-electric theodolite image is introduced. From the both aspects of infrared background pre-processing and infrared target detection, the research statement in China and foreign countries is summarized, also the key problems and difficulties in this field are pointed.Secondly, the feature of the Photo-electric theodolite infrared image is analyzed; based on this, the traditional filter methods in space domain and frequency domain is studied and the application of each filter method is also analyzed, each filter method is tested based on real theodolite images also and the processing effects are compared.Thirdly, the entropy is introduced to denote the complexity of infrared background, and the correspondence relationship is built up based on a large amount of real theodolite images. Base on the noise depression of median filter technique, a target detection method base on Fourier transform and adaptive Butterworth high pass filter is realized and the effect is verified.Finally, the infrared dim target detection method based on maximizing the energy sum of several frames in sequence using dynamic programming is studied, the algorithm is designed and is tested using 200 frames of simulated images. The results shows that the detection and tracking effect of this algorithm is fine, the rate of error detection and miss detection is low, and this algorithm has a great potential in application. |
PDF Full Text Request