Concerning the problem that the numerical prediction method is complex in modeling, the radar echo extrapolation method is easy to generate cumulative error and the model parameters are difficult to set in plateau area, a method for segmenting rainfall clouds in Tibet was proposed based on the improved DeepLab v3. Firstly, the convolutional layers and residual modules in the coding network were used for down-sampling. Then, the multi-scale sampling module was constructed by using the dilated convolution, and the attention mechanism module was added to extract deep high-dimensional features. Finally, the deonvolutional layers in the decoding network were used to restore the feature map resolution. The proposed method was compared with Google semantic segmentation network DeepLab v3 and other models on the validation set. The experimental results show that the method has better segmentation performance and generalization ability, has the rainfall cloud segmented more accurately, and the Mean intersection over union (Miou) reached 0.95, which is 15.54 percentage points higher than that of the original DeepLab v3. On small targets and unbalanced datasets, rainfall clouds can be segmented more accurately by this method, so that the proposed method can provide a reference for the rain cloud monitoring and early warning.

The semi-automatic road extraction method needs more artificial participation and is time-consuming, and its accuracy of road extraction is low. In order to solve the problems, a new method of road extraction from multi-source high resolution remote sensing image based on Fully Convolutional neural Network (FCN) was proposed. Firstly, the GF-2 and World View high resolution remote sensing images were divided into small pieces, the images containing roads were classified by Convolutional Neural Network (CNN). Then, the Canny operator was used to extract the edge feature information of road. Finally, RGB, Gray and ground truth were combined and put into the FCN model for training, and the existing FCN model was extended to a new FCN model with multi-satellite source input and multi-feature source input. The Shigatse region of Tibet was chosen as the research area. The experimental results show that, the proposed method can achieve the extraction precision of 99.2% in the road extraction from high resolution remote sensing images, and effectively reduce the time needed for extraction.

In the background of geological data, in order to more accurately and objectively assess the susceptibility of debris flow, a model of regional debris flow susceptibility assessment based on neural network was proposed, and the accuracy of the model was improved by using Mean Impact Value (MIV) algorithm, Genetic Algorithm (GA) and Borderline-SMOTE (Synthetic Minority Oversampling TEchnique) algorithm. Borderline-SMOTE algorithm was used to deal with the classification problem of imbalanced dataset in the preprocessing phase. Afterwards, a neural network was used to fit the non-linear relationship between the main indicators and the degree of proneness, and genetic algorithm was used to improve the fitting speed. Finally, MIV algorithm was combined to quantify the correlation between indicators and proneness. The middle and upper reaches of the Yarlung Zangbo River was selected as the study area. The experimental results show that the model can effectively reduce the overfitting of imbalanced datasets, optimize the original input dimension, and greatly improve the fitting speed. Using AUC (Area Under the Curve) metric to test the evaluation results, the classification accuracy of test set reached 97.95%, indicated that the model can provide reference for assessing the degree of debris flow proneness in the study area under imbalanced datasets.