An acquiring platform of near-ground remote sensing images is developed to collect visible image and NIR image synchronously and then get the reflectance information from crop images. The platform contains: a 2-CCD multi-spectral camera, two collecting boxes, a field computer with two gigabit net interfaces and an image processing system. First, a true color image of plants is processed with 2G-R-B gray and median filtering algorithm, and the Otsu algorithm is used to separate plants from ground. Then the binary image is returned back to original R gray image and the average gray value of R image is acquired. Finally, using the linear model between reflectance and gray value the reflectance of crop in red waveband is calculated. Test results show that the acquiring platform works well and there exists remarkable correlation between results by above image processing method and that of ASD spectrometer. It provides the theoretical basis for crop status monitoring with digital image processing.