UAV Image Classification for Land Cover Mapping: A CNN-Based Approach in IoT Networks

Remote sensing and land cover mapping have been greatly aided by the advent of unmanned aerial vehicles (UAVs) due to the superior efficiency and detail of the images they provide. To categorize UAV-captured images for land cover mapping, this research introduces a Convolutional Neural Network (CNN)-based technique embedded in Internet of Things (IoT) network architecture. The benefits of UAV imaging and the IoT are combined in the suggested strategy to provide precise and timely land cover categorization. The UAV images used to train the CNN model represent various environments, from forested regions to urban centers to farmland. Due to its capacity for automating visual information extraction, the CNN architecture is well suited to the challenging task of land cover categorization. Connectivity to IoT networks also ensures uninterrupted data transfer and real-time decision-making between UAVs and ground stations. Applications such as urban planning, environmental monitoring, and disaster management greatly benefits from regularly updated land cover maps, which this makes possible. The experimental findings suggest that the proposed CNN-based technique is successful, demonstrating high classification accuracy and resilience across various land cover types. Furthermore, the IoT-enabled system's real-time capabilities are assessed, proving its fitness in applications requiring up-to-the-minute mapping of changing land covers.