图像处理与BP神经网络在矿井火灾隐患识别中的应用

Application of image processing and BP neural network on mine fire potential hazard identification

矿井火灾在矿井工作过程中比较常见,为了保护矿井安全,减少其对煤矿开采工作的影响,提出了一种图像处理技术与BP神经网络相结合的矿井火灾隐患快速识别方法。实验设置了5个场景,共20组不同的图像,把20组不同实验场景图像经过预处理降噪,之后分别通过BP神经网络、SVM支持向量机和K-means聚类3种不同方法进行分析。结果显示,BP神经网络处理后的正确率最高,达到95%。

Mine fire is a common danger in the process of work. In order to protect the mine safety and reduce the fire on coal mining , a fast ident if ic at ion method of mine fire hazard is proposed in this paper, which is based on the combination of image processing technology and BP neural ne twork. The experiments set up 5 scenes. There are 20 groups of different images. The 20 groups of different images can reduce noise by pretreatment. By BP neural network, SVM support vector machine and K-means clustering the data is analyzed. The results show that the correct rate of BP neural network after the treatment reaches 95 %.