Automated grain boundary detection using the level set method

In this paper, we detect the grain boundary by using a novel automated algorithm which is based on the level set method (the LSM). The LSM starts with only one initial closed curve which can be arbitrary in both shape and location in the microphotograph. The curve splits and merges repeatedly until it reaches the boundaries of the objects. The evolution depends on the structure of the image. This method requires simple and feasible constraints which ensure the exclusion of the objects unsuitable for reliable measurements and strain analysis. The LSM has the following three advantages: (1) with a single image as input, segmentation is automated without human intervention, (2) the grain boundary detected by the LSM is a closed curve, and (3) simple post-processing maintains the shape of the boundary. The segmentation and strain analysis by the LSM are demonstrated and the results are compared with those from hand-drawn method.