The present work aims to develop an automated solution for analysis of dimple tensile fracture surfaces. An image processing based technique has been proposed to extract the voids from the fractographic images of dimple fracture surfaces and also to analyse the void features. A two-stage process has been developed to extract the closed contour of the voids. At the first stage a global thresholding is followed by a sequence of processing, like thinning and morphological closing. In the second stage refinement process has been carried out on the large voids based on local characteristics. The proposed technique has been applied on the scanning electron fractographs of tensile fracture surfaces obtained by tensile testing of austenitic stainless steel at four different strain rates. It has been found that the size and density of extracted voids bear systematic relationship with tensile properties of the investigated steel. The average circular diameter with yield strength and tensile strength follows linear relationship. The number density of large voids is found to decrease with increase of strain rate and it also decreases with increase of average circular diameter in a linear manner. Similar nature of variation of tensile ductility has been observed both with strain rate and average circular diameter.

目前的工作旨在开发一种用于分析凹坑拉伸断裂表面的自动化解决方案。已经提出了一种基于图像处理的技术来从凹坑断口的断口图像中提取空隙并分析空隙特征。已经开发了一个两阶段过程来提取空隙的闭合轮廓。在第一阶段，全局阈值处理之后是一系列处理，如细化和形态闭合。第二阶段根据局部特征对大空隙进行了细化处理。所提出的技术已应用于通过在四种不同应变速率下对奥氏体不锈钢进行拉伸试验而获得的拉伸断口表面的扫描电子断口图。已经发现，提取的空隙的大小和密度与所研究钢的拉伸性能具有系统关系。平均圆直径与屈服强度和抗拉强度呈线性关系。发现大空隙的数量密度随着应变速率的增加而减小，并且随着平均圆直径的增加而呈线性减小。在应变速率和平均圆直径方面都观察到了拉伸延展性变化的相似性质。发现大空隙的数量密度随着应变速率的增加而减小，并且随着平均圆直径的增加而呈线性减小。在应变速率和平均圆直径方面都观察到了拉伸延展性变化的相似性质。发现大空隙的数量密度随着应变速率的增加而减小，并且随着平均圆直径的增加而呈线性减小。在应变速率和平均圆直径方面都观察到了拉伸延展性变化的相似性质。