Follow-up Study, Based On Genetic Algorithm For Image

As a genetic algorithm to solve the problem of efficient parallel global search method, with its inherent robustness, parallel and adaptive so that it is very suitable for large-scale search space optimization, has been widely applied in many disciplines and engineering fields. In computer vision and image processing applications fields, it is also growing attention, for the image tracking to provide a new and effective method.The subject of the study mainly provides the Image Processor of TV seeker with a fast, stable image tracking algorithm. A high-resolution image can represent the details of the target, which makes tracking process more stable and reliable. But at the same time this image needs quite a long processing time for image tracking algorithm, and failed to meet the requirements of application of high real-time.Through the deep research and comparison on running mechanism of the genetic algorithm, in allusion to the relevant search strategy of tracking, propose a genetic algorithm based on the relevant match tracking algorithm. The algorithm mainly uses the non-search mechanism of the genetic algorithm to quickly search the best feature of match position among the source image and the target image, shortens the processing time; uses template-updating, and achieves rapid and stable tracking purposes. In order to maintain the diversity of the population, the algorithm uses stochastic and symmetrical approach to generate initial population. Encoding mechanism uses Gray code, is useful to achieve crossover and mutation operator; In order to adapt to the best of the individual as far as possible, the reservation to the next generation groups, select operator to choose the optimal preservation strategy and gambling round choice of combining the choice of strategy; cross uses a uniform cross-strategy, using a variation of the basic-adaptive mutation operator.Programs are all compiled in the Windows XP by VC++ 6.0.