This paper introduces CM-YOLOv8, a lightweight object detection algorithm specifically designed for coal mine fully mechanized mining faces. The algorithm addresses the challenges of limited computational resources and high recognition accuracy demands in coal mining environments. It introduces adaptive predefined anchor boxes tailored to the coal mining dataset to enhance detection performance and employs an L1 norm-based pruning method to significantly reduce model computation and parameter volume without compromising accuracy. The proposed algorithm is validated on the DsLMF+ dataset, achieving a 40% compression rate with less than a 1% drop in accuracy. Comparative analysis with other algorithms demonstrates its efficiency and practicality in coal mining scenarios. The experiments confirm that CM-YOLOv8 reduces computational requirements and model size while maintaining high accuracy, making it suitable for real-time and resource-constrained environments in coal mining.This paper introduces CM-YOLOv8, a lightweight object detection algorithm specifically designed for coal mine fully mechanized mining faces. The algorithm addresses the challenges of limited computational resources and high recognition accuracy demands in coal mining environments. It introduces adaptive predefined anchor boxes tailored to the coal mining dataset to enhance detection performance and employs an L1 norm-based pruning method to significantly reduce model computation and parameter volume without compromising accuracy. The proposed algorithm is validated on the DsLMF+ dataset, achieving a 40% compression rate with less than a 1% drop in accuracy. Comparative analysis with other algorithms demonstrates its efficiency and practicality in coal mining scenarios. The experiments confirm that CM-YOLOv8 reduces computational requirements and model size while maintaining high accuracy, making it suitable for real-time and resource-constrained environments in coal mining.