A method for obtaining a foot region from time-sequential images of foot pressure for gait guidance

A method to analyze gait based on the foot pressure distribution for several steps performed in a natural state would be useful for analyzing gait attributes and walking motion. Extracting each foot region from the foot pressure distribution is a problem in this situation. In this paper, the authors propose a method to measure the position of the foot region by focusing on distinguishing between when the center path of the foot pressure distribution when walking approximates a sine wave or follows a straight line with respect to the walking direction, and to automatically extract the foot region based on grouping the regions around the estimated positions. Based on the results of evaluation experiments, fast, accurate extractions were confirmed for various gait patterns used when analyzing gaits in rehabilitation therapy. The extraction rate was 100% for natural gait in healthy people, and about 98% overall for people with gait impairment. This rate required roughly 1 second of processing. In addition, for gait with shoes, the extraction rate was 100% for natural gait, and 90% when subjects were walking irregularly. This shows that the authors' method can also be applied to the field of crowd dynamics if one person is walking alone, as when giving gait guidance to a healthy person. © 2004 Wiley Periodicals, Inc. Electron Comm Jpn Pt 3, 87(10): 15–26, 2004; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ecjc.20104