Multi-task driving safety state identification method based on visual attention mechanism of driver

The invention discloses a multi-task driving safety state identification method based on visual attention mechanism of a driver. The multi-task driving safety state identification method based on the visual attention mechanism of the driver is aimed to overcome the disadvantage of single parameter of the visual attention mechanism of the driver. According to the multi-task driving safety state identification method based on the visual attention mechanism of the driver, a multi-task driving safety state identification index system is built through gathering and extracting the visual attention parameters of the driver in the normal driving and multi-task driving, weight coefficients of identification indexes are confirmed based on expert analysis and fuzzy network analysis, and a multi-task driving safety state level identification model is constructed according to the weight coefficients. The multi-task driving safety state identification method based on the visual attention mechanism of the driver realizes to precisely identify multi-task driving safety states.