A new image binary segmentation algorithm based on frequency domain

In this paper, a new image binary segmentation algorithm in frequency domain is proposed. After analysing the Fourier transformation, Gaussian kernel function is modified that a new parameter named compression factor is added into, a binary result of the image can be obtained easily and quickly through controlling this new parameter's range. Experimental results show that the proposed algorithm can get satisfactory binary segmentations with a low computational complexity, and the results are better than some traditional binary segmentation algorithms.