Particle measurement by image processing and analysis

To characterize particle size and shape by using image technique, the first step is to delineate parlicle, and then to measure each delineated particle in a right way. For the two steps, there are a number of algorithms have been developed by previous researchers, but there is no standard. This paper describes a genetic algorithm using for delineation of solid particles. It is a new study for particle images. We present the image segmentation system which incorporates a genetic algorithm to adapt the segmentation process to changes in image characteristics caused by variable environmental conditions. After particle delineation, we studied a new algorithm for particle measurement, which is rotation-invariantly, reproducibly, simply, and reasonably. The two algorithms were tested by a number of particle images, they work well in our cases.