<![CDATA[ASugarcane plantations play an important role in food security and the national economy. However, problems such as empty areas or suboptimal planting distances still frequently occur and affect land productivity. This study aims to develop an automatic detection model for empty spaces between sugarcane plants using the YOLOv5 method. The data used consists of 1,000 digital images of sugarcane fields obtained via drone from PT Perkebunan Nusantara III (Persero). The research method follows the AI project cycle, which includes problem scoping, data acquisition, data exploration, modeling, evaluation, and deployment. The labeling of empty areas was performed using Roboflow, and model training was conducted on Google Colab. The model was evaluated using metrics such as IoU, precision, recall, F1-score, and accuracy. The best results showed that the model achieved an accuracy of 35.37%, precision of 0.4774, recall of 0.5771, and F1-score of 0.5225. Additionally, the model results were applied to an interactive dashboard based on Streamlit to facilitate visualization and decision-making in the field. This study demonstrates that YOLOv5 has potential in assisting with the detection of empty areas in sugarcane fields; however, the model's performance can still be improved through data optimization and model parameter tuning.]]>
Copyrights © 2025
