Superpixel Based Hierarchical Segmentation for Color Image

• DOI: 10.1587/transinf.2020edl8025
• Publication Date: 2020-09-30T22:32:42Z
• Authors: Chong WU; Le ZHANG; Houwang ZHANG; Hong YAN
• Keywords: Robustness; Segmentation-based object categorization; Benchmark (surveying); Region growing

In this letter, we propose a hierarchical segmentation (HS) method for color images, which can not only maintain the accuracy, but also ensure good speed. our method, HS adopts fuzzy simple linear iterative clustering (Fuzzy SLIC) to obtain an over-segmentation result. Then, uses fast C-means (FFCM) produce rough result based on superpixels. Finally, takes non-iterative K-means using priority queue (KPQ) refine validation experiments, tested and compared it with state-of-the-art image methods Berkeley (BSD500) benchmark under different types of noise. The experiment results show that outperforms techniques in terms speed robustness.