Deep learning based smart wide area irrigation system

Wide-area irrigations face challenges such as excessive water consumption, heterogeneous exposure to sunlight, and the inability to ensure fair irrigation across the entire surface. As a solution to these problems, it is suggested to customize irrigation strategies according to regional characteristics and utilize modern technologies such as deep learning and image segmentation for efficient use of water resources. The research anticipates that it will provide both water savings and solutions to problems related to heterogeneous plant development due to irrigation in high-water-consuming areas such as golf courses, stadium turfs, and urban landscapes. Therefore, deep learning is expected to be effective in determining plant water needs and predicting irrigation requirements. Simultaneously, the goal is to make the skills of an irrigation-specialized professional accessible to everyone through an artificial intelligence model trained with data obtained from imagebased data collection methods such as live camera images or field photography. The advantages of image-based data collection methods over sensor-based systems include easier applicability, lower cost, and no operational and maintenance costs.