Optimization of Color Dominance Factor by Greedy Algorithm for Leaves and Fruit Segmentation of Tomato Plants

The cultivation of agricultural products destined for human consumption has played a crucial role in the advancement of societies. Consequently, the cultivation of land is facilitated by the integration of various technological advancements into tillage processes. The implementation of different areas of research and technologies have created what is called precision agriculture. In general, precision agriculture systems take information from different sources, which in most cases requires a process of filtering the information. When the source of the information is images, the filtering process in many cases requires a segmentation process to label the pixels of interest. In this work, we present a proposed Greedy Algorithm that seeks to highlight the color dominance present in the leaves and fruits of tomato plants to perform a segmentation process using two \(\alpha \) parameters to optimize. Applying the \(\alpha _{1}=3.2\) and \(\alpha _{2}=2.6\) factors that emphasize color dominance obtained by the metaheuristic algorithm, the performance metrics Accuracy, Precision, Recall, F1-Score and IoU were used, achieving an average of \(86\%\) in fruit segmentation and \(91\%\) in leave segmentation.