针对煤矿生产中输送胶带撒煤问题,采用基于机器人视觉平台的深度学习技术,使用图像分割算法精确定位带式输送机及撒煤区域。通过分析分割特征图和图像分类算法,可以有效判断是否存在撒煤现象。该系统的设计与实践,验证了其在实际应用中的有效性,显著提高了撒煤现象的检测精度,为选煤厂生产提供了安全保障。

To address the monitoring issue of coal spillage in coal mining operations, this study employs deep learning technology based on a robotic vision platform, using image segmentation algorithms to accurately locate the belt conveyor and coal spillage areas. The application results indicate that by analyzing segmented feature maps and employing image classification algorithms, it is possible to effectively determine the presence of coal spillage. The significance of this research lies in its detailed exploration of the system's design and implementation, validating its effectiveness in practical applications. Experimental results show that this system significantly improves detection accuracy for coal spillage, providing safety assurance for coal preparation plant operations.

               
                    1 带式输送机撒煤监测系统
2 理论背景
2.1 图像增强
(1)平移：
(2)旋转：
(3)平移：
2.2 算法总体架构
2.2.1 分割模块
(1)U-Net网络结构概述。
(2)引入ResNet思想及C3结构。
(3)改进SPPF结构。
2.2.2 分类模块
2.2.3 损失函数
2.2.4 总体结构
3 实验与结果分析
4 结 语