提出一种基于图像的快速、准确测量景物近红外波段反射率的反演方法。采用CMOS 相机对同一光照条件下的待测地物与标准反射率板进行测量，采集近红外波段图像，在校正CMOS 相机入瞳处辐照度与图像灰度关系结果的基础上，建立景物近红外反射率与图像灰度的关系，进而反演出对应景物的近红外波段反射率。结果表明，本文方法与地物光谱仪测量结果的不确定度在10%以内。此外，本文还根据景物近红外图像灰度的分布情况分析了背景、测量方位角等因素对地物反射率测量的影响，为进一步精确测量及相关的近红外场景仿真奠定了基础。

This study develops a new image-based retrieval method to measure the near infrared reflectance of objects in a scene rapidly and accurately. The measurement system uses a CMOS camera to capture images of objects and a standard reflectance board under the same light condition in near infrared bands. Based on the corrected relationship between the pupil irradiance of the CMOS camera and the image gray level, the relationship between the image gray level and the near infrared reflectance of scene is established; the near infrared reflectance of the corresponding scene is then found. The results show that the uncertainty of the proposed measurement method is within 10% of the results derived from strictly accurate measurements. Moreover, the influence of the background, measuring the azimuth and other factors on the reflectance measurement, is analyzed according to the gray level distribution of the near infrared scenery images; the proposed method has laid a foundation for the further accurate measurement and near infrared scene simulation.