Morphology Based Sediment Particle Image Binarization Algorithm Research

Abstract

The diversity and complexity of sediment particle images, has become a constraint major bottleneck problem of establishing the river sediment image segmentation, using sediment particle images with similar particle images in other areas can take advantage of same-sex classes of image processing to propose a morphology-based Otsu binarization algorithm, the use of fast binary algorithm to get the initial search reference, on this basis to define a smaller range as the initial search of morphology-based Otsu binarization algorithm, then image segmentation. Experiments show that the new algorithm to maintain the speed and good noise suppression ability, at the same time this algorithm can better retain the original image characteristics, the algorithm execution speed has been greatly improved. Images have good results after binary.