Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device

Apllication of Polar-Mellin Transformation in Optical Joint Transform Correlation Target Recognition

Author LuoShiShang
Tutor ChangShengLi;LiXiuJian
School National University of Defense Science and Technology
Course Physics
Keywords Information Optics Target Recognition Opto-electronic Hybrid Systems Optical Joint Transform Correlator Spatial Light Modulator Distortion Invariant Recognition Polar-Mellin Transform Real-time Performance
CLC TP391.41
Type Master's thesis
Year 2010
Downloads 42
Quotes 0
Download Dissertation

Opto-electronic hybrid target recognition system is used to analyze, judge and classify the targets, by using computers, optical systems and sensors etc. Its basic principle is from the sensory functions of human visual system. It can be widely used in massive image processing, unmanned spacecraft docking, blind land of airplanes, target search and tracking of laser weapon system and so on. Without pre-fabricating holographic filter, the opto-electronic hybrid target recognition system, which is composed of optical joint transform correlator (JTC) and electronic processing unit, can be applied in autonomous target recognition and real-time tracking. In addition to the integration design of optical structures and mechanical and electrical system, the key issue of the application of opto-electronic hybrid target recognition system is its recognition success rate and speed, that is, the realization of target distortion invariant recognition and the high real-time performance of the system.In this paper, an opto-electronic hybrid target recognition system which can realize all-optical Fourier image processing is designed. It is based on 1-f optical JTC and utilizes liquid crystal spatial light modulator (SLM) and CMOS image detectors. The article focuses on analyzing the distortion invariant recognition and real-time recognition of optical JTC.The method of Polar-Mellin transform applied to achieve the distortion invariant recognition of optical JTC, based on lots of numerical simulation and comparison of advantages and disadvantages with other methods. Experiments show that applying Polar-Mellin transform method in image preprocessing can significantly improve the distortion invariant recognition performance of optical JTC.Analysis shows that, under the condition of the good performance of aided electronic processing system, the real-time character of the opto-electronic hybrid target recognition system mainly depends on the the display characteristics and speed of SLM, the speed of image detectors and the work cooperation in time between SLM and detectors. We also find that the performance of SLM for the real-time of recognition system is more critical, according to the development and performance of image detectors. Hence, experiments are designed to test the dynamic display characteristics of liquid crystal SLM. The refresh time and the effective frequency of SLM, obtained from these experiments, are useful for the next optimized design of system with higher real-time perfomance.In chapter 4, according to the theoretical simulation and experimental results, the key factors limiting the speed of optical JTC are analyzed, and the method to improve the distortion invariant recognition performance of optical JTC is proposed. Moreover, a blue print of the integrated system with combination of Field Programmable Gate Array and Digital Signal Processor is provided for improving the real-time performance. The results of this paper provide some advices for the application design of opto-electronic hybrid target recognition system based on optical JTC, and they are useful to exploit the performance of each component of JTC and improve the system performance.

Related Dissertations
More Dissertations