Detection and Counting of the Number of Cocoa Fruits on Trees Using UAV

Abstract

This research aims to recognize and count the number of cocoa pods on trees using an Unmanned Aerial Vehicle (UAV). The main challenge in this research is to estimate the yield of stacked cocoa fruits and count the number of fruits on the trees. Therefore, this research proposes an algorithm that accurately estimates cocoa fruit yield. Image processing is done through many stages: image segmentation, boundary determination, shape analysis, and overlap analysis. The proposed algorithm can recognize and calculate the number of cocoa pods on a tree accurately through overlap analysis between the pods found. The results show that the detection accuracy after applying overlap analysis reaches 94.5%.