Temporal segmentation of environmental map based on changing rate of activity areas and improvement of tracking method using LRF

This paper presents improved methods of tracking moving objects by using laser range finders located in the environment and of segmenting environmental maps weighted by the degree of human activity. The accuracy of our previously proposed method of detecting individuals needing route guidance service, which involves the use of an autonomous mobile robot, depends on the accuracy of pedestrian tracking. Therefore, we herein propose a method that can track a pedestrian even if he or she cannot be directly detected due to occlusion. In addition, since environmental maps depict the human activity histories in particular areas, which vary depending upon the environmental situation, such histories should be divided according to environmental situation to enable more accurate analysis of human activity areas. Thus, we also propose a method of determining the times at which environmental maps should be divided, which is based on calculating the average rate of change in each grid section. Two experiments were conducted to evaluate the usefulness of the proposed methods, and the results are presented herein.