Geological disasters, such as landslides, are vital in terms of life and property preservation and requires timely and accurate monitoring to adequately react to them and respond appropriately. Such catastrophes are to be monitored and acted on by advanced sensor technology that need advanced sensing technologies to operate in-person to detect and track them. This work utilizes satellite remote sensing data with several environmental and geological variables to produce high-precision landslide influence factors. These are the foundations of advanced sensors and monitoring techniques in geological disaster response. The RK-OBGRNet, an integrated recurrence and monitoring model for landslides which uses remote sensing images, is designed to detect and monitor landslides. The preprocessing methods for satellite images are z-score normalization, Fourier Transform, FT which improves the quality of the satellite images by eliminating noise and making data synchronized to the analyzer. It utilizes feature extraction such as principal component analysis to identify characteristics of landslides that dominate their presence. The impact of model parameters on landslide detection accuracy has been analyzed. The performance of RK-OBGRNet is better compared to the RNN-Autoencoder and Cascade R-CNN with improved accuracy of 95%, precision of 94%, recall of 93%, and F1-Score of 92%. The results indicate that RK-OBGRNet achieves better and demonstrates its effectiveness for high-precision landslide detection. This research provides valuable insights for enhancing geological disaster monitoring and emergency response. It offers a reference for applying high-precision sensing technologies to early warning systems and rapid response strategies in the management of landslide hazards.
