Automatic characterization of fracture surfaces of AISI 304LN stainless steel using image texture analysis

Texture analyses methods incorporating three-dimensional fractal analysis using box-counting, grey level co-occurrence matrix (GLCM) technique and run length statistical (RLS) analysis have been carried out on tensile fractographs of AISI 304LN austenitic stainless steel for automatic characterization of fracture surfaces. The tensile tests have been carried out at five different strain rates (0.0001, 0.001, 0.01, 0.1 and 1 s−1). The three above mentioned methods, namely, fractal analysis using box-counting, GLCM and RLS analysis are compared in terms of accuracy and computational time and amongst them the run length analysis shows the best result. Eight texture descriptors from the three texture analyses could be extracted to correlate with the observed mechanical properties. Long run emphasis (LRE) and long run high grey level emphasis (LRHGE) depict better correlation among the eight descriptors in this investigation. The results also reveal systematic variation of image texture properties with strain rate.

► Image based texture analysis as a potential tool to characterize fractographs.
► Fractal analysis, GLCM and run length analysis have been carried out.
► The results show systematic change in image properties with variation in mechanical properties.
► The concept of machine vision has been highlighted.