Boundary Roughness Analysis of Skin Lesions using Local Fractals and Wavelet Transforms
Boundary roughness of skin lesions is of clinicalsignificance for early detection of malignant melanomas. Anintegrated approach of local fractals and multi-level waveletanalysis is proposed in the paper. Local fractal profile of a lesioncontour is generated to describe boundary roughness while a treestructured wavelet packet is further processed to give behaviorsof the local fractals at different frequency bands. Featuresextracted at each sub-bands are selected using Hausdorffdistance for identification of skin tumors in order to reducedimension in feature space. Experimental results show thatfeatures chosen are dominating clues to discriminate skintumors.
Ma Li Huang Kejie Yan Jianqiang Kaihua Wu L.Zhu
School of automation, Hangzhou Dianzi University, Hangzhou, 310018, China Oracle Corporation, 500 Oracle Parkway Redwood Shores, CA 94065, USA
国际会议
成都
英文
1-4
2010-06-18(万方平台首次上网日期,不代表论文的发表时间)