Research on the porosity of concrete sand and stone based on reduced dimension Otsu algorithm

摘要

This paper uses the reduced dimension Otsu fast threshold segmentation algorithm to deal with the concrete sand and stone images. The total area of the pore is calculated by the difference of the grey value, and then the porosity is obtained which can be used to find cement ratio demanded in the concrete mixing process. Compared to the traditional two dimensional Otsu segmentation algorithm, this algorithm decomposes the traditional algorithm into two one-dimensional Otsu algorithms and integrates within and between clusters variance to construct a new threshold discriminant function. With dimensions reduced, the speed and the segmentation effect of the proposed algorithm are better than former works.