Ceramic microscopic image processing based on fast discrete curvelet transform

The properties of ceramic materials are strongly dependent on their microstructure. Ceramic microscopic image processing may be divided into two main procedures: image preprocessing for noise reduction to clarify the image and image segmentation for locating and detecting the objects of interest curvelet transform is a new extension to wavelet transform in two dimensions. Curvelet overcomes the limitation of wavelet in analyzing signals with dimension higher than 1D because it has the character of anisotropy. The fast discrete curvelet transform theory makes it understood and implemented more easily. A novel image denoising method is proposed in this paper. The nonlinear hyperbolic tangential function is selected as curvelet thresholding function. This denoising method is used to ceramic image processing. Firstly, the noisy ceramic image is denoised by the novel curvelet transform thresholding denoising method and then watershed algorithm is applied to the segmentation. Finally, the grain size distribution can be obtained from segmentation image. It has been proved that this method is effective for the ceramic microscopic image.