Level set algorithm based on discrete cosine transform

A new level set algorithm based on discrete cosine transform (DCT) is proposed in order to solve the problem that the traditional level set can only deal with the image with additive noise. For each pixel in the original image, the two dimensional DCT is computed in its local neighborhood. And the denoised image derived from the local DCT coefficients is inspired by the characteristics of the Lee filter. Using the decreasing function of the denoised image gradient as the stop item, the level set function is evolved, and then the final segmentation is produced. Finally, the experimental results show that the developed algorithm could segment the image well while suppressing the multiplicative noise.