摘要
振动筛筛面物料的疏密状态可以作为判断上层旋流器设备堵塞的辅助手段。采用图像处理的方法,使用振动筛筛面网格直线作为检测特征,以支持向量机分类器作为判别方法,设计了一种振动筛物料状态检测算法,利用计算机自动检测辅助人工监控,以及时发现堵塞情况,并减少疏漏。本文首先对总体算法思路进行了阐述;采用开运算与Canny边缘检测强化图像中的直线特征;然后使用基于累计概率霍夫变换的直线检测提取振动筛筛面上的特征直线;最后利用支持向量机作为分类器对提取到的特征进行准确分类。实验结果表明,本文方法能够有效地识别图像中振动筛筛面上物料状态,识别准确率达到97.38%。
The density sparsity of material in the screen of the vibrating screen can be used as an auxiliary mean for judging the blockage of the cyclone. The image processing method is adopted, the vibrating screen mesh lines are used as the detection feature, and the SVM method is utilized to design a vibrating screen material detection algorithm. Computer automatic detection is used to assist manual monitoring, aiming to find congestion and reduce omissions. In this paper, the overall algorithm is described, and the idea of the algorithm is clarified. The opening operation and Canny edge detection in the algorithm play a role in enhancing features. Then the line detection based on the cumulative probability Hough transform is used to extract feature lines on the sieve surface. Finally, the SVM (Support Vector Machine) is utilized as a classifier to accurately classify the extracted features. The experimental results show that the method can effectively identify the presence and absence of materials on the vibrating screen surface, and the recognition accuracy reaches 97.38%.
出处
《图像与信号处理》
2020年第1期57-64,共8页
Journal of Image and Signal Processing
