Lattice Boltzmann Model based 3D Image Segmentation
Yu Chen, Dongxiang Lu · IOP Conference Series Materials Science and Engineering · 2019
Image segmentation plays a very important role in three-dimensional volume data processing. In three-dimensional segmentation, the curve evolution method based on geometric active contour model and level set method is one of the most widely concerned methods. Among these methods, the CV model is a hot research topic. Based on the Munfor-Shah model, CV model is proposed by Chan and Vese. Because of the huge amount of data in three-dimensional images, the computation of solving CV model is very large, which consumes a lot of time. In this study, a 3D image segmentation algorithm based on Lattice Boltzmann model is proposed. The experiment shows our method is effective. The algorithm is simple and efficient. More importantly, the LB model has natural parallelism. Especially, our method can be implemented on massively parallel image processing platform, for example, FPGAs, DSPs and GPUs.