The fuel cell components segmentation and visualization based on CT image

This paper focuses on the image segmentation and three-dimensional visualization based on the CT images of fuel cell. According to the analysis of a variety of traditional edge detection algorithms, Sobel operator and Canny operator are more appropriate on edge detection of the fuel cell tomographic images, and the results of applying them on edge detection of the images are also compared. By analyzing the characteristics of the fuel cell tomographic image grayscale of a particular environment and improving Canny operator, a new method ---- image segementation algorithm for edge detection based on multi-threshold, is designed and implemented, which is also used on two-dimensional image segmentation. Additionally, in order to facilitate a more intuitive observing the fuel cell internal components, this article uses Ray Casting, a commonly used direct volume rendering algorithm of three-dimensional visualization, to rebuild the fuel cell, and applies the multi-threshold image segmentation method on the three-dimensional images to distinguish display the components of liquid water an gas.