Fig. 1. Flow charts of automatic detection of molybdenum mine. (a) X-ray sorting equipment; (b) image detection and control process

Fig. 2. X-ray transmission images of molybdenum ore. (a) Rock; (b) mine

Fig. 3. Enhanced YOLOv10s algorithm tailored for molybdenum ore characteristics

Fig. 4. Comparison of C2f module and C2f-CloAtt module

Fig. 5. Structure of FocalModulation

Fig. 6. Module structure of Dysample

Fig. 7. Point sampling set based on dynamic range factor

Fig. 8. Manual labeling of image judgments

Fig. 9. Ablation study line chart

Fig. 10. Chaos matrix

Fig. 11. Training process of YOLOv10s_pro. (a) Changes in losses during the training process; (b) changes in various indicators during the training process

Fig. 12. Visualization of the detection results in the ablation experiments. (a) Annotated data; (b) detection results of YOLOv10s; (c) detection results of YOLOv10s+C2f-CloAtt; (d) detection results of YOLOv10s+FocalModulation; (e) detection results of YOLOv10s+Dysample; (f) detection results of YOLOv10s+GIoU; (g) detection results of YOLOv10s_pro