Auto Detector for Huanglongbing Citrus Greening Disease using YOLOV7

The disease known as Huanglongbing (HLB), which is frequently seen in citrus plants, is a significant factor in the decreased yield. Because the symptoms have only recently become apparent and can only be identified by someone with subjective knowledge, this disease still poses a threat among citrus diseases. By employing photos of citrus leaves and fruits, YOLOV7-based citrus greening disease diagnosis gets around this problem. Utilizing images of healthy citrus fruits and leaves, typical HLB symptoms, and irrelevant HLB symptoms, three sample datasets are created. When evaluated against these three sets, YOLO V7 did well, earning an overall F1 score of 92%. Then sample set 2 is tested once more under various circumstances using various backgrounds for the images, and it is discovered that the background had little impact on the YOLO V7 prediction.