Vickers Hardness Indentation Segmentation Based On Wavelet Texture Analysis

Hardness of detected material characterizes elasticity, plasticity, strength, toughness, wear resistance and the overall performance of other physical quantities. Vickers hardness is one of the commonly used criteria. The size of hardness is calculated by the indentation area of material surface indirectly. Wavelet multi-resolution texture analysis and clustering are combinated to obtain Vickers hardness indentation contour in this article. And then the indentation area is obtained.Texture feature is used to extracted indentation contour from the perspective of texture segmentation in this paper, which meets the effectiveness requirements of Vickers hardness indentation segmentation and improves range and robustness of image processing algorithm application. Feature extraction and classification is the two main issues of indentation texture segmentation. In this paper, wavelet analysis and clustering as the main tools are used to study the two aspects texture segmentation deeply:(1) A redundant discrete wavelet transform is proposed to extract texture features of indentation based on analyzing advantages and disadvantages of pyramid wavelet transform and discrete wavelet frame transform respectively. And the mean or variance is used as "energy measure" to get stable dimensional texture features of the indentation, which provides a guarantee for texture classification. Texture features based on quartering smoothing algorithm is used to avoid the obvious error brought along the border region when we use variance to calculate texture energy measure of pixels and to solve the boundary effect appearing in indentation texture computing in this paper.(2) The improved k-means clustering algorithm which reduces K-means clustering algorithm running time is used to do texture image segmentation of Vickers hardness indentation. Indentation image is divided into a number of limited areas with different texture features after indentation texture classification and indentation contour is segmented fast and effectivelyThe simulation results indicated that the algorithm is high validity and stability which provides a new idea of Vickers indentation test.