Abstract: There is a strict requirement on the uniformity of bamboo color when bamboo is used as decorative material.A new intelligent classification method on bamboo is presented replacing artificial classification.It uses CIE1976 L*a*b* color space to measure and analyze the surface color of bamboo because L*a*b* space has characteristics of isometry and high resolving power.Before that，the color feature should be converted to L*a*b* space from RGB space.Then the suitable features are extracted to distinguish bamboo and BP neural network is used to classify bamboo in high accuracy.

Key words: L*a*b* color space, neural network, bamboo classification, uniform color space

摘要： 竹片作为装饰用材对颜色的统一性要求严格。利用CIE1976 L*a*b*颜色空间的等距性和高分辨率的特点，将竹片的RGB颜色特征转换成L*a*b*空间颜色特征进行分析，选取能恰当表征竹片颜色的特征参数，再结合BP神经网络分类器对竹片进行细分类。实验表明，分类正确率可达96％，满足工业上的要求，解决了人工目测分类的各种问题。

关键词: L*a*b*颜色空间, 神经网络, 竹片分类, 均匀颜色空间