摘要
基于机器视觉的浮选过程监控方法已经被广泛应用于浮选过程中,泡沫表面纹理特征是过程监控的关键视觉特征之一。当前静态纹理特征只能从空间维度描述图像特征,在时间维度上刻画图像序列的内在变化特性存在不足,不能准确反映浮选泡沫浮选过程动态特性。提出了基于复杂网络时空特性的泡沫图像序列动态纹理特征方法。通过将每帧图像的像素点映射到网络各节点,利用邻接矩阵建立复杂网络模型和网络权值动态演化反应不同时刻的图像特征,基于复杂网络时空特性提取泡沫图像序列的动态纹理特征。结合实际生产数据进行仿真验证,实验结果表明该方法可准确识别浮选动态状况,为浮选生产过程的实时调节提供重要的指导信息。
Methods for monitoring flotation processes based on machine vision has been widely used, which surface texture feature of froths is one of the key visual parameters in process monitoring. Static texture features can only describe images in space dimensions but do not well describe inherent variation characteristics of the image sequence in time dimension, so they can not accurately reflect dynamic characteristics of froths in flotation process. An extraction and analysis method for dynamic texture features of flotation froth images was proposed on a basis of space-time characteristics of complex networks. After pixels of each image were mapped into nodes of complex networks, a complex network model was established by adjacent matrix. The image characteristics at different time were described by network-weighted dynamic evolution and dynamic texture characteristics of image sequences were obtained by the space-time characteristics of complex networks. Simulation results with actual production data showed that the method could accurately identify flotation dynamic conditions and provide guidance for instant regulation of flotation process.
出处
《化工学报》
EI
CAS
CSCD
北大核心
2017年第3期1023-1031,共9页
CIESC Journal
基金
国家自然科学基金项目(61473319)
国家自然科学基金创新研究群体项目(61321003)
中南大学创新驱动计划项目(2016CX014)~~
关键词
浮选
纹理特征
复杂网络
动态建模
过程控制
flotation
texture features
complex network
dynamic modeling
process control
