Detecting Defect in Central Pivot Irrigation System using YOLOv7 Algorithms

Abstract

Central Pivot Irrigation System (CPIS) is a significant method for intelligent irrigation used in global agriculture. It is used to cultivate important crops like wheat, which contribute to global food security. However, the CPIS encounters technical issues that result in malfunctions in its automatic control system, leading to damage to the primary pipes and towers. This can result in significant material losses for farmers and their crops. Repairing the system is also a time-consuming process. To address this issue, a study was conducted using YOLO algorithm which contains several models, and this study used the Yolov7 models to detect defects in the CPIS machine accurately. The study used a dataset gathered from agricultural areas in Salah al-Din Governorate. The models were used to determine whether the CPIS was in a safe or dangerous state. The RGB color system with Yolov7x achieved a 95% detection rate with an accuracy, F1-score, and precision values of 0.798 and 0.954. Similarly, Yolov7x achieved a 93% detection rate with accuracy, F1-score, and precision values of 0.77 and 0.932 in the Grayscale color system. Based on the outcome, this study can conclude that the yolov7x model with RGB Color System accurately detects the CPIS in both its safe and dangerous states. This makes them useful in real-time systems for CPIS defect monitoring and control.