Remote fruit fly detection using computer vision and machine learning-based electronic trap

• DOI: 10.3389/fpls.2023.1241576
• Publication Date: 2023-10-10T06:14:02Z
• Authors: Miguel Molina-Rotger; Alejandro Morán; Miguel Angel Miranda; Bartomeu Alorda-L...
• Keywords: Trap (plumbing); On the fly; Raspberry Pi

Introduction Intelligent monitoring systems must be put in place to practice precision agriculture. In this context, computer vision and artificial intelligence techniques can applied monitor prevent pests, such as that of the olive fly. These are a tool discover patterns abnormalities data, which helps early detection pests prompt administration corrective measures. However, there significant challenges due lack data apply state art Deep Learning techniques. Methods This article examines classification fly using Random Forest Support Vector Machine algorithms, well their application an electronic trap version based on Raspberry Pi B+ board. Results The combination two methods is suggested increase accuracy results while working with small training set. Combining both for yields 89.1%, increases 94.5% SVM 91.9% RF when comparing all species other insects. Discussion research reports successful implementation ML system detection, providing valuable insights benefits. opportunities IoT devices image opens new possibilities, emphasizing significance optimizing resource usage enhancing privacy protection. As grows by increasing number traps, more will available. Therefore, it holds potential further enhance learning from multiple systems, making promising effective sustainable population management.