Application of image segmentation of aero-engine based on genetic algorithm and region growth

摘要

Endoscope detection segmentation is important for the detection of numerous faults and plays an important role in the automatic endoscope detecting systems. In this paper, we propose a new image segmentation algorithm introducing threshold selecting based on the genetic algorithm into the region growth. Genetic Algorithm is powerful in optimizing the feature vector through declining useless features and finding optimal ones. The GA is used to choose an optimal threshold to complete the seed generation. Gray-level and texture are used to merge regions with similar characteristics. Experimental results show that the proposed segmentation effect of the method is ideal, and the efficiency is improved.