基于PNN的退化交通标志图像的识别算法研究

李伦波; 马广富

doi:10.3724/SP.J.1146.2007.01638

基于PNN的退化交通标志图像的识别算法研究

doi: 10.3724/SP.J.1146.2007.01638

基金项目:

高等学校博士学科点专项科研基金(20050213010)和国家自然科学基金(60674101)资助课题

计量
- 文章访问数: 4049
- HTML全文浏览量: 74
- PDF下载量: 746
- 被引次数: 0
出版历程
- 收稿日期: 2007-10-16
- 修回日期: 2008-01-30
- 刊出日期: 2008-07-19

Identification of Degraded Traffic Sign Symbols Using PNN

摘要

摘要: 为了识别退化的交通标志图像，该文采用一种新的特征提取算法。该算法在处理图像退化问题时，采用模糊-仿射联合不变矩直接提取图像的特征，从而避免了需要较大计算量的图像复原处理过程。针对各阶模糊-仿射联合不变矩数量级差异较大问题，提出一种数量级标准化算法。在深入分析PNN与K-means聚类算法的基础上，提出采用全局K-均值算法优化设计概率神经网络分类器，并将其用于交通标志图像的分类识别。仿真结果表明：模糊-仿射联合不变矩是一种有效的处理退化交通标志图像的方法，所设计的概率神经网络分类器不仅具有精简的结构而且具有较好的推广性能。
- 模式识别; 概率神经网络; 交通标志; 模糊-仿射联合不变矩; 全局K-均值算法
Abstract: A novel feature extraction algorithm is presented for the recognition of traffic sign symbols undergoing degradations in this paper. In order to cope with the degradations, the Combined Blur-Affine Invariants (CBAIs) are adopted to extract the features of traffic sign symbols without any restorations which usually need a great amount of computations. A new magnitude normalization method is proposed for the great differences of magnitude of combined blur-affine invariants. Under the deep discussion of PNN and K-means algorithm, a probabilistic neural network classifier is designed using global K-means algorithm and applied to the classification of degraded traffic signs. The simulation results indicate that CBAIs are efficient for the feature extraction of degraded images, and the designed network is not only parsimonious but also has better generalization performance.

HTML全文

参考文献(1)

De la Escalera A and Salichs M. Road traffic sign detectionand classification[J].IEEE Trans. on Industrial Electronics.1997, 44(6):848-859[2]Gavrila D. Multi-feature hierarchical template matchingusing distance transforms. IEEE 14th InternationalConference on Pattern Recognition, Brisbane, Australia,1998: 439-444.[3]Miura J, Kanda T, and Shirai Y. An active vision system forreal-time traffic sign recognition. Proceedings of IEEEIntelligent Transportation Systems, Dearborn, MI, USA,2000: 52-57.[4]Fleyeh H. Shadow and highlight invariant color segmentationalgorithm for traffic signs. IEEE Conference on Cyberneticsand Intelligent Systems, Bangkok, Thailand, 2006: 1-7.[5]Douville P. Real-time classification of traffic signs[J].Real-TimeImaging.2000, 6(3):185-193[6]Nguwi Yok-Yen and Z Kouzani Abbas. Detection andclassification of road signs in natural environments[J].NeuralComputing Applications.2007, DOI:10-[7]Gavrila D. Traffic Sign Recognition Revisited. Proc. of the21st DAGM Symposium, Bonn, Germany, 1999: 86-93.[8]Cyganek B. Circular road signs recognition with affinemoment invariants and the probabilistic neural classifier.Lecture Notes in Computer Science 4432, Springer, Berlin,2007: 508-516.[9]Suk T and Flusser J. Combined blur and affine momentinvariants and their use in pattern recognition. PatternRecognition, 2003, 36(12): 2895-2907.[10]Hu M K. Visual pattern by moment invariants. IEEE IRETrans. on Information Theory, 1962, 8(1): 179-187.[11]Reiss T H. The revised fundamental theorem of momentinvariants[J].IEEE Trans. on Pattern Analysis and MachineIntelligence.1991, 13(8):830-834[12]Flusser J and Suk T. Pattern recognition by affine momentinvariants[J].Pattern Recognition.1993, 26(1):167-174[13]Specht D F. Probabilistic neural networks for classification,mapping or associative memory. IEEE InternationalConference on Neural Networks, Piscataway, New Jersey,USA, 1988: 525-532.[14]Masters T. Practical Neural Network Recipes in C++. SanDiego: Academic Press Inc, 1993: 201-222.[15]Likas A, Vlassis N, and Verbeek J J. The global K-meansclustering algorithm[J].Pattern Recognition.2003, 36(2):451-461[16]Hansen P, Ngai E, and Cheung B K, et al.. Analysis of globalK-means, an incremental heuristic for minimum sum-ofsquaresclustering[J].Journal of Classification.2005, 22(2):287-310

施引文献

资源附件(0)

访问统计