Research on Automatic Location of Seed Points in Ultrasound Breast Tumor Images Based on Fuzzy Logic Algorithm
Tianyu Zhao, Hang Dai · 2021 International Conference on Aviation Safety and Information Technology · 2021
Breast cancer is the second leading cause of cancer death in women, which seriously endangers women's health. Breast tumor segmentation is one of the most critical and difficult tasks. In order to realize the automatic and rapid positioning of seed points and meet the needs of real-time online image segmentation, this paper presents an automatic positioning method of seed points in ultrasound breast tumor images based on fuzzy logic algorithm. The fuzzy logic is innovatively introduced into neural network, which is combined with the method of recalibrating the importance of characteristic elements in attention mechanism. By defining a reasonable fuzzy membership function, different diffusion coefficients are adopted for different pixel gradients. Automatic reference points are obtained by using the automatic reference point selection method based on texture features of image gray level co-occurrence matrix. On this basis, the seed points are iteratively obtained by Mean−Shift algorithm, which is the tumor area. Simulation results show that the proposed method is superior to most existing fuzzy closed-value segmentation methods.