Comparative Analysis Of Deep Learning Models For Semantic Segmentation Of Uav Images

Semantic segmentation of UAV images is vital for applications like agriculture, urban planning, and disaster management. This study compares two deep learning models, Deeplabv3+ and SegNet, using metrics such as Kappa coefficient, pixel accuracy, mIoU, recall, F1 score, and precision. Both models achieved over 90% accuracy in identifying Arecanut, coconut, fallow land, and buildings, with Deeplabv3+ outperforming SegNet in accuracy and training time. The findings highlight the strengths and weaknesses of each model, aiding in the selection of an optimal model for specific UAV data.