IoT Drone Camera for a Paddy Crop Health Detector with RGB Comparison

This paper proposes the system of paddy crop health detector using drone camera. Indonesia is an agricultural country that has very large agricultural land, where every plant health monitoring activity is done manually. However, applying technological developments in land monitoring activities will shorten time and increase work efficiency. In this paper a drone with a raspberry pi camera has been used to capture several images of rice fields from several areas. The image data is processed into a digital leaf color chart (LCC) through the process of image acquisition, RGB color extraction, and k-Nearest Neighbor (k-NN) classification. The data has been compared with the real LCC, which is a reference to the health color of rice plants. The paddy fields that are used as the research material are 25 days after planting. The result shows that the precision of the method is 88.89%, the recall is 93.02%, the accuracy is 98.22%, and the specificity is 98.77%.