Jixie qiangdu (Jan 2016)
RECOGNITION METHOD OF METAL FRACTURE IMAGES BASED ON GROUPLET TRANSFORM AND KERNEL PRINCIPAL COMPONENT ANALYSIS
Abstract
Grouplet transform is a new directional wavelet. This wavelet can be transformed at any time and space,and adaptively change the basis according to image texture. Therefore Grouplet transform has a good ability of sparse representation.Here,Grouplet transform is introduced into the metal fracture images,and combined with the Kernel Principal Component Analysis（ KPCA）,a new recognition method of metal fracture images based on Grouplet-KPCA is proposed. At the same time,the proposed method is compared with the wavelet-KPCA recognition method. The experimental results show that the proposed method can overcome the information of finite directions only obtained by the wavelet-KPCA recognition method,and can have a satisfactory recognition rate. Compared with Grouplet entropy,Grouplet kurtosis is more sensitive to the texture change of metal fracture and suitable for feature extraction of metal fracture. Therefore the recognition method based on Grouplet kurtosis-KPCA have better recognition rate than the recognition method based on Grouplet entropy-KPCA.
