Artificial Intelligence to Identify the Nectar Site: A Deep Learning Approach

The rising standard of living has led people to pay more attention to the importance of natural foods. Honey as a nutritional product is beneficial to human health. At present, honey production mainly relies on nectar sites, however, due to temperature, humidity, the flowering period and plant growth status of nectar sites are difficult to be accurately grasped. In this paper, we propose a deep-learning nectar site identification model, which can distinguish nectar sites from non-nectar sites in real time at the macro level, and can accurately determine whether a field monopoly and nectar plants are in flowering stage at the micro level using satellite remote sensing data. It is important for grasping the distribution of nectar sites in real time and promoting the development of smart agriculture.