Flood detection regions model using whale-crow search optimization based deep convolutional neural network

The advancements in satellite images have attracted more attention in the field of flood detection. Flood detection is an important task for planning actions during emergencies, but the major obstacle is to detect the flooded regions using satellite images. This paper design a model named Whale-crow search algorithm based deep convolutional neural network (W-CSA DCNN) model for flood detection. The proposed model undergoes four steps namely pre-processing, segmentation, feature extraction, and classification. At first, a satellite image is given to pre-processing for extracting noise and artifacts from the input image. Then, the pre-processed image is subjected to the feature extraction process for extracting the features based on vegetation indices. The obtained features are then used in the segmentation process, which is done using Kernel Fuzzy Auto regressive (KFAR) model. Once the segments is obtained, then the segments are given to the classification, which is performed using the DCNN, which is trained optimally using the proposed W-CSA that is obtained from the combination of Crow search Algorithm (CSA) and whale optimization algorithm(WOA).