摘要
提出了一种基于局部形态和彩色特征的回转窑烧结状态识别方法。使用SIFT描述回转窑图像的局部形态特征,将彩色的回转窑图像的三个分量都转化为八个等级,用局部三维直方图来描述局部的颜色和亮度特征,将两种特征融合得到局部形态和彩色特征。使用词袋(Bag-of-Words)模型表示图像并利用神经网络分类器实现对烧结状态的识别。实验结果表明,基于局部形态和彩色特征的识别方法能够获得较高的识别精度。
A new method based on local shape and color features is presented to recognize the sintering states in rotary kiln. Local shape features are represented by SIFT descriptor and local color and brightness features are described by the local 3-dimensional histogram after the three components of the color rotary kiln image are converted into eight levels,then the two kinds of features obtained are fused into the local shape and color features. The Bag-of-Words model is used to represent the image and then the neural network classifier is applied to recognize the sintering state. Experimental results show that the recognition method that is based on local shape and color features can obtain higher recognition accuracy.
出处
《计算机工程与应用》
CSCD
北大核心
2015年第18期194-198,共5页
Computer Engineering and Applications
基金
湖南省自然科学基金(No.13JJ3050)
国家自然科学基金(No.61203016
No.61174050)
中央高校基本科研业务费资助
关键词
烧结状态
特征描述
彩色特征
回转窑
sintering state
feature description
color feature
rotary kiln