Volume : VIII, Issue : VIII, August - 2019

APPLICATION OF OTSU SEGMENTATION ALGORITHM BASED ON SIMILARITY IMPROVEMENT IN FABRIC DEFECTS

Jiao Cuijiao, Yang Lianhe

Abstract :

In order to improve the segmentation result of faic defect image, an Otsu segmentation algorithm based on similarity improvement is proposed. Firstly, by compå the median filtering and Gaussian filtering, the average filtering with better filtering effect is used to filter the faic defect image, which reduces the influence of noise on the segmentation result, and effectively suppresses the interference of the background texture of the faic defect image; Then, by the histogram analysis of the faic defect image, for the classical Otsu segmentation algorithm has no better segmentation result of the single–peak type faic defect image, the threshold is adjusted by calculating the similarity between the segmented defect image and the original defect image, enhancing the ability of the Otsu segmentation algorithm to adapt the single–peak type images. Experiments show that the Otsu segmentation algorithm based on the similarity improvement has better segmentation results for the single–peak type faic defect image, and can also obtain better segmentation results for the bimodal type faic defect image, which has certain reference value in the faic defect research

Keywords :

Article: Download PDF   DOI : 10.36106/ijsr  

Cite This Article:

APPLICATION OF OTSU SEGMENTATION ALGORITHM BASED ON SIMILARITY IMPROVEMENT IN FABRIC DEFECTS, Jiao Cuijiao, Yang Lianhe INTERNATIONAL JOURNAL OF SCIENTIFIC RESEARCH : Volume-8 | Issue-8 | August-2019

Number of Downloads : 128

References :