摘要
为了识别退化的交通标志图像,提出了一种新的分类算法。该算法在处理图像的退化问题时,采用模糊—仿射不变距直接提取图像的特征而不需要图像的清晰化处理;在利用模糊—仿射不变距提取图像特征的基础上,采用递归正交最小二乘算法设计了一种新的径向基概率神经网络分类器。仿真结果表明:模糊—仿射不变距是一种有效的处理退化的交通标志图像的方法,所设计的径向基概率神经网络分类器不仅具有精简的结构,而且,具有较好分类和推广性能。
A novel classification method is presented for recognizing traffic sign symbols undergoing image degradations . In order to cope with degradations, the classifier uses the combined blur-affine invariants（CBAIs） of traffic sign symbols as the feature vectors which allow to recognize objects in the degraded scene without any restoration. A radial basis probabilistic neural network （RBPNN） is designed with recursive orthogonal least algorithm（ROLSA） and applied to the classification of degraded traffic signs. The simulation results indicate that CBAIs are efficient to the feature extraction of degraded images and the classification and generalization performance of the RBPNN classifier with the reduced structure are good.
出处
《传感器与微系统》
CSCD
北大核心
2007年第8期43-47,共5页
Transducer and Microsystem Technologies
关键词
交通标志
径向基概率神经网络
模糊-仿射不变距
递归正交最小二乘法
traffic sign
radial basis probabilistic neural networks （ RBPNN ）
combined blur-affine invariants （CBAIs）
recursive orthogonal least algorithm（ROLSA）
