Research on Automatic Landslide Identification Based on Faster R-CNN and Mask R-CNN
Based on the images from Gaofen-1 satellite,taking the first section of the Three Gorges reservoir as an example,160 landslide samples are visually interpreted and divided into training and validation samples according to the ratio of 9∶1.Then,landslide automatic identification models are constructed using Mask R-CNN and Faster R-CNN algorithms.To further compare and analyze the performance of the two models under different sample ratios,calculations are performed using sample ratios of 8∶2,7∶3,and 6∶4.The results show that the recognition results based on Mask R-CNN model are better than Faster R-CNN in three indicators,including precision,recall and Fl score.Mo-reover,cross-validation proves that the performance of Mask R-CNN model is more stable.
deep learninglandslide identificationMask R-CNNFaster R-CNNcross-validation
万方数据
维普
