Identifying the maturity level of coconuts using deep learning algorithms

Identifying the maturity level of the coconuts is challenging for coconut cultivators. Up to now, many coconut cultivators in India are manually determining the maturity level. It is difficult for a person can climb the coconut tree and identify the maturity level of the coconuts. Few farmers started the deployment of robots to harvest the mature coconuts from the tree. In this process, the robot determines the maturity of coconuts using image processing techniques. This article discusses an efficient and accurate method for determining the maturity level of coconuts. The authors proposed a dataset for detecting raw and ripened coconuts in this work. Further, to train the dataset, the authors used various deep learning algorithms such as YOLO-V5s, YOLO-V4 Tiny, and Mobilenet-SSD, respectively. Finally, the results related to different deep learning algorithms have been compared in terms of Precision, Recall, and F1 score.