煤粒图像识别系统的设计与实现

Design and implementation of coal particle image identification system

设计开发了一个用于现场自动控制的煤粒图像识别系统,并介绍了该系统的硬件系统设计和软件系统设计。利用该系统可以提取煤粒图像的一些特征参数作为神经网络的输入,煤质指标作为训练目标,训练出所需要的神经网络并保存下来。利用这个网络可以精确预测当前煤质指标,从而可为现场的煤质指标控制提供可靠数据依据。

A coal particle image identification system applied to the site automatic control was designed and developed.The partial codes of the MATLAB programming for the system ware provided.The system could be applied to extract some feature parameters of the coal particle images as the input of the neural network.Taking the coal quality index as the training target,the required neural network could be obtained from the training and could be stored.The neural network could be applied to accurately predict the present coal quality index and thus could provide the reliable data basis to the site control of the coal quality index.