Artificial Intelligence based Satellite Image Classification Using Advanced Deep Learning Technique

Identification of areas using remote sensing areas is required for the detection of climatic changes and control, precisely for disaster monitoring and related important environment monitoring. Satellites are almost universally employed to identify disasters on the ground and offer some benefits in imaging the Earth. The advanced image processing technique helps to get clear information about the Earth’s surface. Flood detection has become more convenient in the recent past. The proposed research system has developed the capacity to detect areas covered by floods using deep learning techniques. In this research work, the Advanced Segmentation Network (AdSegNet) system is used for implementing an image processing technique for disaster detection. The decision-making is made using AdSegNet through which the disaster analysis can be extracted more efficiently. Correcting strong overlapping of pixels is advantageous to AdSegNet for the accessibility of image characteristics and false predictions that impact the assessment of the flood zone. The AdSegNet includes established classification originating from the images from satellites before training and testing. The accuracy of the proposed AdSegNet system is about 96.1%.