RESEARCH ON THRESHOLD SEGMENTATION ALGORITHM AND ITS APPLICATION ON INFRARED SMALL TARGET DETECTION ALGORITHM

For infrared dim target detection, proposed an improved Otsu algorithm, advantages in the search for the optimal solution using genetic algorithms, use class variance between the image pixels as a judgment function, realized the improved algorithm, reached the class variance detection threshold segmentation algorithm improvements. Infrared image processing by real infrared video sampling, the proposed algorithm with traditional the Otsu algorithm simulation comparison, experimental results show that the proposed algorithm guaranteed under the premise of the detection accuracy of the segmentation algorithm, to the improvement of the performance of real-time processing of traditional algorithms, excellent algorithm program with a value of engineering practice.