| 退化交通标志图像的RBPNN分类算法研究 |
| |
| 引用本文： | 丁淑艳,宋婀娜,李伦波. 退化交通标志图像的RBPNN分类算法研究[J]. 计算机仿真, 2010, 27(1): 281-284,304 |
| |
| 作者姓名： | 丁淑艳 宋婀娜 李伦波 |
| |
| 作者单位： | 1. 黑龙江科技学院电气与信息工程学院,黑龙江,哈尔滨,150027
2. 哈尔滨工业大学控制科学与工程系,黑龙江,哈尔滨,150001 |
| |
| 摘 要： | 为了识别退化的交通标志图像,采用模糊-仿射联合不变矩直接提取图像的特征,并针对各阶模糊-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,避免了需要较大计算量的图像复原处理过程。同时在深入研究径向基概率神经网络的基础上,采用全局K-均值算法优化其网络结构,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,所设计的径向基概率神经网络分类器不仅具有精简的结构而且有较好分类精度和推广性能。
|
| 关 键 词： | 径向基概率神经网络 交通标志 模糊-仿射联合不变矩 |
Classification of Degraded Traffic Sign Symbols Using RBPNN |
| |
| DING Shu-yan,SONG E-nuo,LI Lun-bo. Classification of Degraded Traffic Sign Symbols Using RBPNN[J]. Computer Simulation, 2010, 27(1): 281-284,304 |
| |
| Authors: | DING Shu-yan SONG E-nuo LI Lun-bo |
| |
| Affiliation: | 1.College of Electrical & Information Engineering/a>;Heilongjiang Institute of Science and Technology/a>;Harbin Heilongjiang 150027/a>;China/a>;2.Dept.of Control Science and Engineering/a>;Harbin Institute of Technology/a>;Harbin Heilongjiang 150001/a>;China |
| |
| Abstract: | For the recognition of degraded traffic sign symbols,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 computation.A new magnitude normalization method is proposed for the great differences of magnitude of combined blur-affine invariants.By deeply discussing the radial basis probabilistic neural network(RBPNN),a novel structure optimization algorithm for RBPNN is proposed using global K-means a... |
| |
| Keywords: | Radial basis probabilistic neural networks (RBPNN) Traffic sign Combined blur-affine invariants (CBAIs) |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
