篇名 | Hybrid Intelligent Data Fusion Approach to Collision Warning Information Extraction |
---|---|
卷期 | 13:2 |
作者 | Bao Rong Chang 、 Hsiu-Fen Tsai |
頁次 | 120-129 |
關鍵字 | Vision analysis 、 Collision warning 、 Embedded dual-platform 、 Data fusion 、 Drowsy driving 、 EI 、 SCI 、 SCIE 、 Scopus |
出刊日期 | 201106 |
Based on vision sensing and dedicated short range communication (DSRC) disseminating car’s GPS information via v2v communication, a fast collision warning response to an imminent crash has been successfully developed. An embedded dual-platform, DaVinci+XScale-NAV270 was employed to achieve line-marking identification, neighboring vehicles detection, and headway measurement. In order to tackle the problem of driving drowsiness in a vehicle, this paper has also considered the specific four factors: fatigue level, breath alcohol concentration, carbon monoxide concentration, and carbon dioxide concentration. To do so, this study applied a data fusion QT-BPNN/ANFIS to fuse heterogeneous data and then infer the precise collision warning signal. The experimental results show that the proposed approach outperforms two alternative well-known systems.