基本信息:
- 专利标题： 一种融合多种特征的高速公路隧道停车事件分级识别方法
- 专利标题（英）：Multi-feature fusion-based classification identification method for parking events in highway tunnel
- 申请号：CN201710010428.8 申请日：2017-01-06
- 公开(公告)号：CN107066929A 公开(公告)日：2017-08-18
- 发明人： 赵敏 , 孙棣华 , 孙健 , 王荣斌 , 唐毅 , 蒋陈虎
- 申请人： 重庆大学 , 重庆首讯科技股份有限公司
- 申请人地址： 重庆市沙坪坝区沙坪坝正街174号;
- 专利权人： 重庆大学,重庆首讯科技股份有限公司
- 当前专利权人： 重庆大学,重庆首讯科技股份有限公司
- 当前专利权人地址： 重庆市沙坪坝区沙坪坝正街174号;
- 代理机构： 北京汇泽知识产权代理有限公司
- 代理人： 武君
- 主分类号： G06K9/00
- IPC分类号： G06K9/00 ; G06K9/32
The invention discloses a multi-feature fusion-based classification identification method for parking events in a highway tunnel. By researching static and dynamic features of parking targets in a tunnel scene, feature difference of actual parking targets and fake parking targets is analyzed, and a multi-feature fusion-based classification identification method for the parking targets in the highway tunnel scene is designed. According to the method, firstly in combination with corrected lane regions of interest, periodic features of a foreground are extracted by utilizing a lane division method and a multi-frame foreground fusion mode and are analyzed and processed, and then whether the parking events occur or not is judged stage by stage by taking the dynamic centroid features as main features and taking the static color and area features as auxiliary features, so that effective identification of the parking targets is realized and the accuracy of detecting the parking events in the highway tunnel at present is improved.
公开/授权文献:
- CN107066929B 一种融合多种特征的高速公路隧道停车事件分级识别方法 公开/授权日：2021-06-08
IPC结构图谱:
| G | 物理 |
| --G06 | 计算；推算；计数 |
| ----G06K | 数据识别；数据表示；记录载体；记录载体的处理 |
| ------G06K9/00 | 用于阅读或识别印刷或书写字符或者用于识别图形，例如，指纹的方法或装置 |