| 基于维纳复原的道路限速交通标志检测 |
| |
| 引用本文： | 乔 敏,孙国强.基于维纳复原的道路限速交通标志检测[J].教育技术导刊,2020,19(5):234-237. |
| |
| 作者姓名： | 乔 敏 孙国强 |
| |
| 作者单位： | 上海理工大学 光电信息与计算机工程学院，上海 200093 |
| |
| 摘 要： | 超速引发的交通事故数量急剧增加，为了提高道路驾驶安全性，提出一种基于维纳复原的道路限速交通标志检测方法。首先，对原始街景进行维纳滤波复原图像预处理；然后，根据道路限速标志牌轮廓形状和颜色等特征，在 HSI 彩色空间进行红色阈值分割、Canny 边缘检测和 Hough 圆检测，对图像中的道路限速交通标志进行检测，定位出可能是道路限速交通标志的区域；最后，根据限速标志的内部特征排除定位出的非限速交通标志区域。实验结果表明，该方法对于我国常见的 6 类限速标志检测正确率达到了 98.76%，符合限速标志检测系统的准确率要求。
|
| 关 键 词： | 限速交通标志 检测 维纳滤波器 二值化 伪目标排除 |
| 收稿时间： | 2019-07-30 |
Speed-limit Traffic Road Signs Detection Based on Wiener Restoration |
| |
| QIAO Min,SUN Guo-qiang.Speed-limit Traffic Road Signs Detection Based on Wiener Restoration[J].Introduction of Educational Technology,2020,19(5):234-237. |
| |
| Authors: | QIAO Min SUN Guo-qiang |
| |
| Institution: | School of Optical-Electrical and Computer Engineering，University of Shanghai for Science and Technology，Shanghai 200093，China |
| |
| Abstract: | With the sharp increase in the number of traffic accidents caused by speeding，in order to improve road driving safety，a road speed limit traffic sign detection method based on Wiener restoration is proposed. First，the original street scene image is subjected to Wiener filtering to restore the image preprocessing；then，according to the shape and color of the road speed limit sign outline，red threshold segmentation，Canny edge detection and Hough circle detection are implemented in the HSI color space. The roadspeed-limiting traffic signs in the image are detected to locate possible areas of road speed-limiting traffic signs. Finally，the non-speed-limiting traffic sign areas located are excluded based on the internal characteristics of the speed-limiting signs. The experimental results show that the accuracy rate of this method for detecting 6 types of speed limit signs in China is 98.76%，which meets the accuracy requirements of speed limit sign detection systems |
| |
| Keywords: | teaching emotion affinity artificial intelligence |
|
| 点击此处可从《教育技术导刊》浏览原始摘要信息 |
| 点击此处可从《教育技术导刊》下载免费的PDF全文 |
