Abstract: Proposes a non-invasive prosthetic vision system via auditory pathway for blind people. The system acquires the surrounding information to which the EBCOT (embedded block coding with optimized truncation) algorithm is used before using the wavelet transform to code ROI (region of interest). Then, the data for image restructuring are obtained via the surrounding information transmission and decoding. The data are scanned in terms of polar coordinates, from which the coordinate information on each and all pixels and their gray levels in an image are mapped into the time, frequency and oscillation amplitude of acoustic signals which are then synthesized together in sine-wave form. The results showed that the image data can be coded and transmitted quicker to acquire more information if using EBCOT algorithm, and the acoustic signals have better recognizability if scanning image data in terms of polar coordinates for mapping.