Abstract: When processed agricultural remote sensing monitoring in the south of China, remote sensing images are often affected by thin cloud. Atmospheric absorption and scattering effect can make the sensor receives ground reflectance differs from the true value, which is the main reason for the decline of remote sensing data. It is necessary to remove thin cloud and fog and make the atmospheric correction. This study used the LandSat-7/ETM+ image, made the cloud removal by BSHTI-VCP method, contrast the result by dark element method, made the FLAASH atmospheric correction with the processed image, analyzed and evaluated the spectral characteristics and NDVI value of typical objects before and after correction. The results showed that the BSHTI-VCP method can lower pixel gray value of visible light and near-infrared wave with 0.3341-0.5476 and 0.0591-0.2512, separately, 0.0529-1.0729 increase in image average gradient, and raise information entropy, too. The BSHTI-VCP method can effectively eliminate the influence of thin cloud and fog on remote sensing data which increased the image quality of cloud cover range. FLAASH atmospheric correction effectively eliminates the effect of atmosphere and can obtain the ground truth surface reflectance, improve the spectral characters of cropland surface, obviously. This study provides theoretical basis for crops remote sensing monitoring further quantitative inversion and information interpretation in the south of China.