|
| 交通路标图像的分割与识别方法研究 |
| Study on separation and identification of traffic signs images |
| |
| DOI: |
| 中文关键词: 路标 HSV颜色空间 分割 尺度不变特征变换 识别 |
| 英文关键词:traffic signs HSV color space segmentation SIFT identification |
| 基金项目: |
|
| 摘要点击次数: 3940 |
| 全文下载次数: 563 |
| 中文摘要: |
| 对交通道路标志进行实时、正确的识别,是车辆自动导航中的一个重要方面。提出一种基于尺度不变特征变换(SIFT)算法的交通路标识别方法。该方法首先根据交通路标的色彩信息,利用HSV颜色空间对输入含有路标的彩色图像直接进行处理,从而快速分割提取目标区域;然后基于SIFT特征将分割提取的目标图像与数据库中原有图像进行特征匹配,以实现交通路标的识别。实验采用MatLab进行仿真,实验证明该方法具有较高的识别精度和速度,在车辆自动导航中具有较高的应用价值。 |
| 英文摘要: |
| To identify the traffic signs correctly and in real time is an important part of vehicle auto navigation. A method of traffic sign identification based on scale invariant feature transform (SIFT) is presented. According to the color of traffic signs, the color images with traffic signs are processed directly based on HSV color space first by this approach, so that the objective region is segmented and extracted quickly. And then features matching of the extracted objective image and the existing image in the database for traffic signs identification based on the SIFT features are done. The experiments do with MatLab simulation, it show that this method has higher identification precision and speed which has a high application value in vehicle auto navigation. |
| HTML 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |