Unmanned aerial vehicle (UAV) near-air remote sensing technology provides the accessibility to monitor the farmland in a rapid and real-time manner. By taking use of the UAV visible light remote sensing platform, here the aerial images of canopy using 26 ramie germplasms were generated and analyzed for the characteristic values using the image processing pipeline. The results showed that HSV color image segmentation can effectively recognize ramie from soil weeds. The variation coefficient at 6 phenotypic traits of 26 ramie resources was 11.00%-52.39%, and the diversity index was 0.62-1.58. The variation coefficients of 15 canopy color and texture traits of 26 ramie resources were distributed between 0.28% and 48.09%, and the diversity index was ranged from 1.25 to 1.54. That indicated a broad phenotypic variation in the tested ramie germplasm resources. Two principal components were identified by principal component analysis of 15 canopy color and texture traits, and the cumulative contribution rate reached 95.10%, which can effectively reflect the main information of each trait.