On the performance analysis of Yolo algorithms to identify the vector of Greening in orange farms

Brazilian citrus farming plays a crucial role in the country’s agricultural economy, boasting significant volumes of citrus fruit production that contribute to both domestic consumption and the export market. However, the advent of Huanglongbing (HLB), commonly known as Greening disease, has introduced substantial challenges to orange juice production, threatening productivity and profitability. This disease is mainly spread by the Asian citrus psyllid (Diaphorina citri), making early and accurate detection of this insect vector essential to mitigate its impact. This study evaluates and compares the performance of state-of-the-art object detection algorithms, YOLOv7, YOLOv8, YOLOv9, YOLOv11, and YOLOv12, in identifying (Diaphorina citri) in orange groves. Experimental results demonstrate that YOLOv9 outperforms other versions based on evaluation metrics, highlighting its potential to improve vector monitoring and control strategies in citrus farming. These findings provide insights into the application of advanced computer vision techniques for agricultural pest management. A key contribution of this work is the ability to identify the disease transmitter in its early stage, where it is not yet capable of contaminating fruits. The model can be integrated with practical implementation aspects, such as its potential use in drone monitoring and intelligent capture systems.