Haptic data compression for rehabilitation databases

Abstract

A rehabilitation database is a concept that utilizes the quantitative data acquired from rehabilitation robots. By applying database techniques to rehabilitation robot data, many applications will become possible. As one example, this paper discusses how to match rehabilitation data based on a dynamic programming method. It is to be anticipated that large amounts of data and long calculation times for searching will be two serious issues for rehabilitation databases. Therefore, this paper proposes a method based on two techniques: feature extraction and nonlinear quantization. Both techniques have the combined features of data compression and good recognition performance. Hence, the matching of compressed data has a high recognition rate, even if the compression ratio is very high. The performance of the proposed method is evaluated through experimental data of 500 trials.