摘要
提出了一种羽毛杆折痕检测方法。将羽毛杆图像转化为一维信号,利用小波变换模极大值和信号奇异点的关系完成折痕位置预判,通过该位置完成子图像提取,减少对羽毛杆遍历检测带来的误判;提出局部Radon变换计算子图像不变矩,通过改变尺度因子获得矩不变量矩阵,利用奇异值分解获得特征量;采用决策级融合方法得到最终识别结果。实验结果表明,该方法有着较低的非折痕误测率和较高的折痕识别率,适用于羽毛杆表面缺陷的自动识别。
A detection method of feather quill crease is proposed. The feather quill image is transformed into one dimensional signal, using the relationship between wavelet transform modulus maxima and singular point, the crease coordinate is prejudged. Subimage is extracted through the coordinate to reduce misiudgeement caused by image traversal. With local Radon transform to extract moment invariants of target subimage, invariants matrix is obtained by changing the scale factor, singular value decomposition is provided to obtain feature invariant for recognition. Finally, the final recognition result of the system is achieved by the fusion of identification results at the decision level. The results show that this suggested method has high recognition rate and is applicable to recognition of feather quill.
出处
《计算机工程与设计》
CSCD
北大核心
2013年第12期4351-4355,共5页
Computer Engineering and Design
基金
广东省科技计划基金项目(2012B020314005)