Abstract
Investigation of situation of riprup is very important to examine the function of the dam used for long period. Among several investigations, the picture of the riprup is very useful to know the particle distribution and existing ratio of classification of rock. However, this is very time consuming and troublesome work when the investigation is carried out by manual procedure. In this paper, it is examined that the procedure is carried out automatically by using the image analysis. The most likelihood method and the maximum posterior probability method are used to judge the difference among rock classifications. The data used for judge is RGB data of each pixel. Moreover, the discrimination method of each rock from the binarized image is newly developed. The proposed method is applied to regular and irregular riprups. The results are compared with the ones from manual procedure. The particle distribution form image analysis is very similar to the ones from manual method. While the existing ratio of classification of rock is a little different, the practical application would be effective.
