2000 Volume 12 Issue 3 Pages 238-243
This paper is concerned with the development of a haptic sensor system that monitors the human skin conditions. The base of the sensor is an aluminum shell, around which a sponge rubber layer, a PVDF film, a protective acetate film and a gauze are stacked in sequence. The sensor is attached onto the tip of an elastic cantilevered beam and pressed against the human skin. It is then moved by hand over the skin to collect the surface morphological features through the sensor. The characteristics of data are extracted by calculating the variance and the distribution of power intensity over several segmented frequency ranges. Each set of data obtained is used as a training set of data for a neural network that distinguishes the skin with atopic dermatitis from the normal healthy skins. Experimental results show that 100% of recognition is achieved by using the constructed network system.