Detection of particle motion using image processing with particular emphasis on rolling motion

Abstract

Image-processing has been used in granular systems for detecting particle positions and motion near optically accessible surfaces like sediment flow and bedload transport. We review the image-processing techniques used for single and multiple particles. To enhance reliability in particle recognition, tools like Canny edge and Hough transform are intensively used. We show exemplarily how they can be applied to detect not only particle positions but also rotatory motion. The different steps are described in detail and the algorithm is applied to different examples, which are discussed in view of the obtained accuracy.