This paper presents an improved version of the YOLOv8 model, named Our-v8, for the detection of coal and gangue. The authors address the lightweight and real-time issues in coal sorting detection by enhancing the YOLOv8 model with several modifications. The dataset includes images of coal and gangue under two different lighting environments (halogen and fluorescent lamps). The Laplacian image enhancement algorithm is introduced to improve the quality of training data, sharpen contours, and boost feature extraction. The CBAM attention mechanism is used to prioritize crucial features, enhancing feature extraction accuracy. Additionally, the EIOU loss function is added to refine box regression, further improving detection accuracy. The experimental results show that Our-v8 achieves excellent performance in detecting coal and gangue in a halogen lamp lighting environment, with a mean average precision (mAP) of 99.5%, FLOPs of 29.7, parameters of 12.8, and a model size of only 22.1 MB. Our-v8 provides accurate location information for coal and gangue, making it suitable for real-time coal sorting applications. Compared to other YOLO series models, Our-v8 demonstrates higher accuracy and faster detection speed while maintaining a lightweight model. The study concludes that Our-v8 is a promising solution for coal and gangue detection, offering high accuracy, speed, and lightweight characteristics.This paper presents an improved version of the YOLOv8 model, named Our-v8, for the detection of coal and gangue. The authors address the lightweight and real-time issues in coal sorting detection by enhancing the YOLOv8 model with several modifications. The dataset includes images of coal and gangue under two different lighting environments (halogen and fluorescent lamps). The Laplacian image enhancement algorithm is introduced to improve the quality of training data, sharpen contours, and boost feature extraction. The CBAM attention mechanism is used to prioritize crucial features, enhancing feature extraction accuracy. Additionally, the EIOU loss function is added to refine box regression, further improving detection accuracy. The experimental results show that Our-v8 achieves excellent performance in detecting coal and gangue in a halogen lamp lighting environment, with a mean average precision (mAP) of 99.5%, FLOPs of 29.7, parameters of 12.8, and a model size of only 22.1 MB. Our-v8 provides accurate location information for coal and gangue, making it suitable for real-time coal sorting applications. Compared to other YOLO series models, Our-v8 demonstrates higher accuracy and faster detection speed while maintaining a lightweight model. The study concludes that Our-v8 is a promising solution for coal and gangue detection, offering high accuracy, speed, and lightweight characteristics.