Research Article
Analysis, Design and Implementation of a Ripe Mango Detection Program in Burkina Faso
@INPROCEEDINGS{10.1007/978-3-031-81573-7_15, author={Moustapha Bikienga and Roland Manegaouind\^{e} Tougma and Souma\~{n}la Ouedraogo}, title={Analysis, Design and Implementation of a Ripe Mango Detection Program in Burkina Faso}, proceedings={Towards new e-Infrastructure and e-Services for Developing Countries. 15th International Conference, AFRICOMM 2023, Bobo-Dioulasso, Burkina Faso, November 23--25, 2023, Proceedings, Part II}, proceedings_a={AFRICOMM PART 2}, year={2025}, month={2}, keywords={Deep learning YOLO algorithm Mangoes Performance metrics Computer vision}, doi={10.1007/978-3-031-81573-7_15} }- Moustapha Bikienga
Roland Manegaouindé Tougma
Soumaïla Ouedraogo
Year: 2025
Analysis, Design and Implementation of a Ripe Mango Detection Program in Burkina Faso
AFRICOMM PART 2
Springer
DOI: 10.1007/978-3-031-81573-7_15
Abstract
The analysis, concept, and implementation of a computer program capable of detecting ripe mangoes are at the core of this study. Traditional methods are often faced with calibration errors. Recently, deep learning has shown promising performance in visually guided agricultural applications. Faced with these constraints, it is necessary to establish an automatic system for robust and efficient detection of mangoes in orchards. In this study, a fast implementation system of a mango detector, distinguishing between ripe and unripe mangoes based on deep learning using the YOLOv5 algorithm, was developed. From a simple photo, the algorithm detects and counts the number of mangoes on a tree. This artificial intelligence system (deep neural network) was trained on a dataset of over 500 annotated mango images. Experimental results show that the algorithm achieves 98% precision, 98% recall, and an F1-score of 98%. This satisfactory precision in mango detection offers significant advantages in terms of efficiency and accuracy compared to traditional methods. However, it should be noted that our system has certain limitations. Nevertheless, our study demonstrates promising results in the field of ripe mango detection.
