IR Thermal Image Segmentation Based on Enhanced Genetic Algorithms and Two-Dimensional Classes Square Error

An enhanced Image segmentation of IR thermal images based on two-dimensional classes square error is discussed. Aimed at the low distinguish ability, low SNR of IR thermal images and very high computation cost of Image segmentation of two-dimensional classes square error, a new image segmentation algorithm based on Chaos-Genetic Algorithms is proposed. The experiments’ results show that, because the grey of point and the average grey of area have been carefully taken into account and the Chaos-Genetic Algorithms has been adopted, the new algorithm can obtain very good image segmentation at a very low computational cost, and the enhanced algorithm is effective and valuable.