On detection of motion by wavelet transform for an acoustic visual aid system

Towards the realization of an acoustic visual aid system, we have made basic studies on a system that transforms images to an array of sound sources. The system can display images of simple static solid bodies and characters through active acquisition of a subject using a head-mounted camera. We show how to display moving bodies, and propose a simple algorithm based on a wavelet transform. This is a frequency analysis method of the output signal of a spatial filter based on a wavelet transform. The algorithm of wavelet-spatial filtering enables us to detect both positions of moving points and their approximate speeds