True-color image of Mars is important for mineral analysis. The working light source of the Mars exploration multi-spectral camera is different from the laboratory calibration light source, so that the commonly used RGB calibration method has a large color difference in correcting the on-orbit images. After analyzing the influence of light source lightness and relative spectral distribution on the camera output values, their influences are considered separately in the calibration. A new color calibration method in the luminance-independent chromaticity space is proposed. The RGB color components of multi-spectral camera are normalized and converted to chromaticity values, and the polynomial calibration model in chromaticity space is compared with the RGB method. The experimental results showed that the average value of CIE DE2000 color differences after color calibration was 4.32, which was 0.96 less than that of RGB method. When the calibration matrix was used to correct the outdoor scene images, the corrected color difference gain because of the change of imaging light source was reduced by 0.54 compared to the RGB method. The color calibration method and the calibration data in this paper provide theoretical support for the color correction of the Mars multi-spectral camera on-orbit images.

火星的真彩色图像对于矿物分析很重要。火星探测多光谱相机的工作光源与实验室校准光源不同，因此常用的RGB校准方法在校正在轨图像时色差较大。在分析了光源亮度和相对光谱分布对相机输出值的影响后，在校准中将分别考虑它们的影响。提出了一种与亮度无关的色度空间中的颜色校正新方法。将多光谱相机的RGB颜色分量归一化并转换为色度值，并将色度空间中的多项式校准模型与RGB方法进行比较。实验结果表明，CIE DE2000色标校正后的色差平均值为4.32，比RGB方法小0.96。当使用校准矩阵校正室外场景图像时，与RGB方法相比，由于成像光源的变化而导致的校正色差增益降低了0.54。本文中的颜色校正方法和校正数据为火星多光谱相机在轨图像的颜色校正提供了理论支持。