Extraction of mechanoluminescent pattern based on afterglow images

A novel technique has been developed to observe the stress distribution by the mechanoluminescent (ML) sensor. The ML materials are able to convert mechanical action to light intensity directly. Dynamic stress distributions on surface of various structure are visualized as patterns of light intensity by the ML paint sensor that is composed of ML micro-particle and binder. This technique has been applied for evaluation of artificial hard tissue such as synthetic femur. It should be noted that ML phenomenon is accompanied with undesirable afterglow which intensity decreases according to time progressing. In this study, a novel extraction method of the ML patterns based on afterglow images is proposed. We assumed uniformity of decreasing function of afterglow intensity. An average pattern of afterglow images provides base pattern of afterglow. Polynomial approximation of dot products between observed images and the base pattern provides component values of afterglow pattern. By subtracting computed afterglow pattern from observed images during some load working, ML patterns are successfully extracted.