Abstract It’s very important to alert drivers about driving environments and possible car crash with other vehicles. Developing real-time automotive driver assistance systems has attracted much attention recently. In order to accomplish such a system, we have tried a camera mounted on a vehicle to capture road scenes for lane and preceding vehicle detection. Images tracking in old times were captured by color webcam. Webcam cannot work in the dark and influenced easily by shadow. Shadow changing of driver’s face cannot avoid when driving on road. We avoid shadow changing with Kinect in the thesis. We get driver’s depth information of head image then processing and tracking. Using the motion of driver’s head as the line of vision for driving safety. Research method: first, cut out background and get foreground by threshold. After morphology and connected component process, using histogram and face feature to judge face then tracking by particle filter and get head motion. Driver’s line of vision is represented by the angle of head rotation. Head rotation changes the length of shoulders. Estimate the angle of head rotation by short shoulder length divided by long shoulder length.
Real-Time Analyzing Driver’s Image for Driving Safety
2019-01-01
5 pages
Article/Chapter (Book)
Electronic Resource
English
European Patent Office | 2016
|Real-time Determination of Driver's Driving Behavior during Car Following
British Library Conference Proceedings | 2015
|Method and Apparatus for Evaluating Driver's Driving by Analyzing Video
European Patent Office | 2024
|Real-time Determination of Driver's Driving Behavior during Car Following
SAE Technical Papers | 2015
|Detection of Driver's Face Orientation for Safety Driving Assistance
British Library Online Contents | 2010
|