Improved Canny Algorithm to Realize Extraction of Wood Texture Features

Abstract

Because in the edge detection the traditional Canny algorithm need to artificially set Gaussian filter parameters and double thresholds when the threshold is lagged, and exist "slope" and "ladder" in non maximum suppression. So that this paper uses the adaptive median filtering RAMF algorithm instead of Gaussian filter, using the improved non maximum suppression to thin the ridge and the Otsu algorithm to replace of the operator with double thresholds to track the boundary. Using the Tamura and statistical methods to extract the texture feature of musical material of the Ruanxian after the edge detection. The results of the experiment show that the improved Canny algorithm is better than the traditional Canny algorithm for edge detection effect; the texture extraction method of the paper can provide theoretical basis for the selection of material for the Ruan.