基于提升小波变换的图像可听化方法  

An Image Sonification Method Based on Lifting Wavelet Transform

作　　者：田亚男[1] 王旭[1] 

机构地区：[1]东北大学信息科学与工程学院,辽宁沈阳110819

出　　处：《东北大学学报（自然科学版）》2011年第2期223-226,共4页Journal of Northeastern University(Natural Science)

基　　金：国家自然科学基金资助项目(60970157)

摘　　要：传统的图像可听化方法多是在空间域选择映射特征,且声音编码维度比较单一,使得盲人学习和使用这种声音的难度较大.针对这一问题,提出了一种基于变换域的图像可听化方法,即对图像小波变换低频系数进行映射,使用频率和音长进行声音编码,将图像表示为电子音乐.实验表明,盲人对这种方法表示的图像具有较高的识别率,并且声音的听感优美,长时间收听不易疲劳.In the conventional image sonification method the mapping features selected in the space domain and the simple sound coding make the blind difficult to understand such a sound.A method based on the transform domain is therefore presented,where the low frequency coefficients of wavelet transform are mapped into the frequency and sound length so as to represent the image in electronic music.Experimental results showed that the blind is able to understand the images mapped into the auditory representation.Moreover,the sound is pleasing to ear and enjoyable.

关 键 词：电子行走辅助 视觉替代 可听化 听觉显示 小波变换 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]