基于BP神经网络的催化剂烧结图像识别  

The Recognition of Catalyst Sintering Image Base on BP Neural Networks

作　　者：李含光[1] 赵晓瑜[1] 

机构地区：[1]南京信息工程大学计算机与软件学院,南京210044

出　　处：《江南大学学报（自然科学版）》2010年第2期181-184,共4页Joural of Jiangnan University (Natural Science Edition)　

摘　　要：为了丰富化工催化剂产品活性的识别方法,通过对化工催化剂使用过程中的图像进行滤波、分割,利用灰度共生矩阵提取部分重要的特征值。提出了用BP神经网络训练成化工催化剂产品活性识别器的新方法。实验结果表明:对化工催化剂产品有效性识别率高,减少企业对催化剂的盲目使用,节约了生产成本。The new method of getting BP neural network trained to identify the chemical catalyst product activity is put forward through the use of chemical catalysts in the process of image filtering and segmentation to enrich the chemical catalyst activity of identification.The process extracted some important features of value by taking advantage of the gray level co-occurrence matrix.The result of experiment shows that the recognition of chemical catalyst effectiveness is high.Misusing chemical catalyst can be reduced as well as production costs can be saved in the enterprise.

关 键 词：神经网络 BP算法 图像识别 催化剂活性 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]