Improved FCM based distance regularization level set algorithm for image segmentation

Considering the traditional contour initialization with high sensitivity to noise limitation in original level-set algorithm, this paper proposes a Distance Regularized Level Set Evolution (DRLSE) evolution model based on the morphological reconstruction FCM algorithm. First, morphological reconstruction (MR) is used to smooth and denoise the image. Under the premise of ensuring anti-noise and image detail preservation, the image is clustered and segmented as the initial contour of DRLSE evolution, and finally the clustering result is segmented by DRLSE model evolution. . The experiment is carried out on multiple images. The segmentation results of the algorithm in this paper are significantly better than the traditional FCM algorithm and DRLSE model, which improves the accuracy of segmentation and is robust to noise.