摘要： 黑色素是人体皮肤中最为重要的一种色素,目前常用的非入侵式估计方法大多以光谱仪为基础,其价格非常高昂,限制了该类方法的应用。为此,提出一种由可见光图像估计黑色素分布的快捷算法。以光学和皮肤生理学为基础,分析皮肤颜色的形成过程并用Elman神经网络模拟该逆模型,从而获得与颜色值对应的黑色素分布。在不同光照及相机拍摄条件下的皮肤图像中进行实验,结果表明,所提算法比其他同类算法的估计结果更准确,可以帮助医学工作者实现非正常黑色素分布的快捷诊断。

关键词: 黑色素, 光学, 皮肤生理学, 逆模型, Elman神经网络

Abstract: Melanin is one of the most important pigments in human skin.At present,the most commonly used nonintrusive estimation methods are based on spectrometer which is very expensive,thus restricting the application of these methods.Therefore,a fast algorithm for estimating the distribution of melanin from visible images is proposed.Based on the principles of optics and skin biophysics,the process of skin color formation is analyzed and the inverse model is simulated by an Elman neural network,so as to obtain the corresponding melanin distribution.The algorithm is tested on skin images with different illumination and camera shooting conditions.Experimental results demonstrate that the proposed algorithm performs more accurately than other similar methods on estimation results,which can help medical workers to diagnose non-normal distribution of melanin quickly.

Key words: melanin, optics, skin biophysics, inverse model, Elman neural network

中图分类号: 

• TP301.6