正负模糊规则系统、极限学习机与图像分类  

Positive and negative fuzzy rule system,extreme learning machine and image classification

作　　者：吴军[1] 王士同[1] 赵鑫[1] 

机构地区：[1]江南大学数字媒体学院,无锡214122

出　　处：《中国图象图形学报》2011年第8期1408-1417,共10页Journal of Image and Graphics

基　　金：国家自然科学基金项目(60704047);国家自然科学基金项目(9082002)

摘　　要：传统的图像分类一般只利用了图像的正规则,忽略了负规则在图像分类中的作用。Nguyen将负规则引入图像分类,提出将正负模糊规则相结合形成正负模糊规则系统,并将其用于遥感图像和自然图像的分类。实验证明,其在图像分类过程中取得了很好的效果。他们提出的前馈神经网络模型在调整权值时利用了梯度下降法,由于步长选择不合理或陷入局部最优从而使训练速度受到了限制。极限学习机(ELM)是一种单隐层前馈神经网络(SLFN)学习算法,具有学习速度快,泛化性能好的优点。本文证明了极限学习机与正负模糊规则系统的实质是等价的,遂将其用于图像分类。实验结果说明了极限学习机能很好的利用正负模糊规则相结合的方法对图像进行分类,实验结果较为理想。The positive fuzzy rules often were used only for image classification in the traditional image classification system, while the negative image classification rules were ignored in effect. Nguyen introduced the negative Fuzzy rules into the image classification, proposed a combination of positive and negative fuzzy rules to form the positive and negative fuzzy rule system, and then applied it to remote sensing image/natural image classification. Their experiments proved that their proposed method has achieved good results. However, since their method was realized using the feed forward neural network model which adjust the weights in the gradient descent, the training speed is very slow. Extreme learning machine （ELM） is a single hidden layer feed forward neural network （SLFN） learning algorithm, which has advantages such as quick learning, good generalization performance. In this paper,it proves that Extreme Learning Machine （ELM） and the positive and negative fuzzy rule system is essentially equivalent, so ELM can be naturally used for image classification. Our experimental results support this claim.

关 键 词：图像分类 正负模糊规则系统 极限学习机 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]