介绍了一种基于DMD的哈达玛变换成像光谱仪的物理模型,由于DMD微镜单元尺寸与探测器像元尺寸不匹配或者其他一些在光学设计和系统集成中的误差,导致这种光谱仪中采集到的编码图像上部分像素点出现的一种光谱混合现象。描述和分析了这些光谱混合像素点的编码原理,分析表明:这种光谱混合像素点无法直接用哈达玛反变换解码。为了准确地解码这些光谱混合像素点,改善复原后光谱图像的质量,提出了一种针对这种光谱混合像素点的解码方法。首先向光谱仪中导入一束激光并充满整个视场,针对编码图像中的光谱混合像素点确定一个解码系数。然后将这个系数带入光谱混合像素点的编码方程,即可解码它的光谱元。实验结果证明该方法是可行的。
A sort of engineering Hadamard transform spectral imager based on DMD was introduced. Because the size of individual micromirrors did not match the pixel size of the detector, or for other unavoidable errors in the optical design and the system assembling, a sort of spectral mixing phenomenon appeared on some pixels of the encoded images acquired from the detector. Their encoding principle was described and analyzed. The analysis demonstrated that these spectral mixing pixels could not be decoded by inverse Hadamard transform directly. To decode these spectral mixing pixels accurately and improve the quality of the decoded spectral images, a decoding method for these spectral mixing pixels of the encoded images was proposed. Firstly, a decoding coefficient for the spectral mixing pixel was gotten by directing a beam of laser into the spectral imager to fill the field of view. Then the coefficient was brought into the encoding equation of the spectral mixing pixel to decode all the spectral elements. The experimental results prove the feasibility of the decoding method.