Publication details
Visual Diagnostics of Vehicle Driver
| Authors | |
|---|---|
| Year of publication | 2008 |
| Type | Article in Proceedings |
| Conference | AiM2008 |
| MU Faculty or unit | |
| Citation | |
| Field | Informatics |
| Keywords | image processing; HMI; fatigue detection; attention |
| Description | The statistics has proven that the major cause of traffic accidents is insufficient attention paid to the traffic situation by the driver. One of the most common causes of decreased attention is fatigue and the probably most serious consequence is a microsleep. Our objective is to propose a robust method for diagnostics of drivers' level of attention by the means of visual recognition; this method is intended to be used with conjunction with other methods (e.g. EEG assessment) to achieve better accuracy. This contribution presents the our concept of "visual diagnostics" and our preliminary experiments with AAM method. We have sought for optimal parameters and training set for the AAM model creation and tested them against certain states that were expected to have significant impact on the recognition results. We have identified several factors that have significant negative impact on the accuracy and proposed guidelines to build the model. We also concluded that the parameters influence the accuracy and the robustness against particular phenomenons depends actually less than the respective training set. The results indicates that the method is suitable for target environment and can reliably provide desired information with sufficient reliability. |
| Related projects: |