Abstract
In this study, the spatial local optimization method was improved to obtain high precision of optical flow for cases in which the object movement changes substantially and a method to trace the loci of moving objects was considered. In the spatial local optimization method, the precision of the optical flow when the object movement changes substantially becomes a problem. Therefore, to make the object movement relatively small, we obtained flow vectors from the image sequence to drop the resolution of the original input image sequence to half the initial resolution. flow vectors were then obtained from the original input image sequence that were smaller than the threshold value. We show that the precision of the optical flow when the object movement changes substantially is improved by this method. Method used to trace the loci of moving objects was demonstrated. We obtained clusters from histograms of flow vectors and pursued each cluster. We show that it is possible to trace moving objects by this method.
Similar content being viewed by others
References
Miike H, Koga K (eds) (1993). Video image processing using a personal computer. Morikita, Tokyo, pp 133–143.
Xu G, Tsuji S (1998) 3-D, vision. Kyoritsu, Tokyo, pp. 111–118
Agui T, Nagao T (1992) Image processing and image recognition. Shokodo, Tokyo, pp 92–98
Author information
Authors and Affiliations
Corresponding author
About this article
Cite this article
Sugisaka, M., Sato, S. Improvement of the precision of the gradient method and object tracking using optical flow. Artificial Life and Robotics 7, 182–184 (2004). https://doi.org/10.1007/BF02471203
Received:
Accepted:
Issue Date:
DOI: https://doi.org/10.1007/BF02471203