11: Detecting twospotted spider mite and mite eggs under strawberry leaf using deep learning

Twospotted spider mite (TSSM: Tetranychus urticae) is one of the most harmful pests for the strawberry plant. TSSM usually feeds on strawberry leaves and affects photosynthesis. Counting of TSSM (larva, protonymph, deutonymph, and adult) and TSSM egg can help growers to apply the right amount of pesticide at the right location at the right time. With the rapid development of deep learning, many object detection methods have been developed in the past several years, including R-CNN, Fast-RCNN, Faster R-CNN, You Only Look Once (YOLO), and SSD. These methods outperformed the traditional computer vision and machine learning methods. Therefore, this study adopted a deep learning-based object detection method in computer vision for TSSM and TSSM egg counting. Eight hundred and seventy-five images were collected using a smartphone with a macro lens. Then, one of the state-of-the-art object detection methods, YOLO, was used to detect TSSM and TSSM egg. The mean average precision (mAP) was 0.62, and the average precision for counting TSSM and TSSM egg was 0.53 and 0.71, respectively. Due to the narrow camera depth of field and different illumination conditions, many of the collected images were blurred at the edge area and contained many shaded areas. The detection performance can be further improved when some pre-processing steps are applied to the images. This study shows that deep learning can effectively detect TSSM and TSSM egg for efficient pest management.