基于独立成分分析法ICA的多光谱重建

构造了一个6通道,8-bit多光谱相机(MSC)模型作为图像采集装置(包括一个三色数码相机和一个滤光片),在多光谱成像技术基础上,通过模拟相机的数码响应,对物体表面的反射光谱进行重建。由数码相机的响应值得到物体的光谱反射比。使用独立成分分析法来进行数据压缩,光谱重建结果分别用CIE1976色差,光谱反射比RMS误差和反射比相对误差进行评价。20例训练样本的平均CIE1976色差为0.1579,平均反射比均方差为0.0099,平均反射比相对误差为0.0107;12例待测样品的平均CIE1976色差为0.7150,平均反射比均方差为0.0309,平均反射比相对误差为0.0312。结果说明,基于独立成分分析法的光谱重建是一种有效的光谱重建方法。

Multi-spectral reconstruction based on the method of independent component analysis (ICA)

A six-channel and eight-bit multi-spectral camera(MSC) was built, which consists of a tri-chromatic digital camera and an optical filter and works as an image-capturing device.It can reconstruct the reflectance of the interest object through simulating the digital count of the camera model based on the multi-spectral imaging technique.That is,the spectral reflectance can be estimated from the digital count.The data are compressed by the method of independent component analysis(ICA),and the reconstruction results are evaluated by the CIE1976 color difference,the reflectance RMS error factor and the reflectance relative error.The average CIE1976 color difference,the average RMS and relative error of the 20 training samples are 0.1579,0.0099 and 0.0107 in turn.The 12 testing samples are 0.7150,0.0309 and 0.0312 in turn.The results show that the method of spectral reconstruction based on ICA is an effective method of spectral reconstruction.