Real-Time Object Detection for the Running Train Based on the Improved YOLO V4 Neural Network

Figure 8

Image after data enhancement. (a) Magnification and brighten, (b) magnification and mirror, (c) magnification and darken, (d) image reduction and translation, (e) image reduction and translation, and (f) mirror, reduction, and translation.