Insect detection using YOLO-based models depends on high-quality image data, but the effects of different imaging devices have not been sufficiently explored. This study compares two imaging methods—a flatbed scanner (Epson Perfection V850 Pro) and a smartphone (Samsung Note 10)—for capturing images of insects on sticky traps. The taxonomic order of the insects was annotated by two entomologists and the images from each device were used to train a YOLOv11 model. The performance was analyzed using mAP, precision and recall and their significance was compared using one-way ANOVA. Our result showed that there were no significant differences between the model trained with scanner and smartphone images. However, the scanner performed significantly better than the smartphone in predicting smaller Psocoptera insects. These results emphasize the importance of using appropriate devices for image acquisition to improve insect detection performance in deep learning applications.