Improved OTSU and adaptive genetic algorithm for infrared image segmentation

Abstract

In order to improve the segmentation result of infrared images, an image segmentation method based on improved OTSU method and improved genetic algorithm is proposed. Firstly, morphological noise reduction is carried out by using morphological weighting adaptive algorithm and then global optimization of OTSU image segmentation function is carried out by using improved genetic algorithm. The method can automatically adjust the genetic control parameters according to the individual fitness and the degree of population dispersion, which can speed up the convergence while maintaining the diversity of the population. Finally, the optimal threshold for image segmentation is obtained, which overcomes the poor convergence, premature and other issues of the traditional genetic algorithm. Experiments show that the threshold range obtained by the method is more stable, the calculation time of the threshold is greatly reduced, and the requirement of real-time image processing can be satisfied.