Research And Application Of Fatigue Driving Detection Algorithm Based On Facial Feature Analysis

In recent years,with the rapid development of the transportation industry,vicious road traffic safety accidents have also risen simultaneously,and it is recognized as the greatest threat and killer of mankind outside the war.Statistics show that there are many incentives for motor vehicle traffic accidents.Among them,fatigue driving accounts for a high proportion,posing a huge threat to traffic safety regulations.The use of modern scientific and technological means to reasonably determine the driver’s fatigue status,to realize real-time detection and alarm,is an effective measure to reasonably prevent traffic accidents,and is also an important subject of current road traffic safety management research.Affected by individual differences among different drivers and complex light changes,occlusions,and interference in actual outdoor driving environments,there are still many technical bottlenecks in the research of driving fatigue detection technology with high real-time and robustness.Aiming at solving the above problems,this paper focuses on the driver’s face detection,face tracking,and fatigue feature extraction in practical application environments,and analyzes the change law of multiple fatigue characteristic parameters under different fatigue states in order to design an efficient reliable fatigue detection system.Based on the driver’s facial visual characteristics,this paper relies on its non-invasiveness,low cost,scalability,and friendliness,accurately extracts relevant fatigue characteristics in real time for parameter analysis to complete the accurate detection and evaluation of fatigue status.The specific research content is as follows:(1)For the complex driving environment,select an appropriate convolutional neural network algorithm for face detection.In order to solve the defect of traditional SSD algorithm detection performance and meet the requirements of practical application scenarios,innovative improvements were made and performance comparison analysis was performed on the self-built face dataset.Experimental results show that the improved SSD algorithm has great advantages in detection performance.(2)Based on the real-time requirements of the fatigue detection system design,a face tracking and prediction algorithm based on Camshift was designed.By integrating with the target detection algorithm,a more efficient extraction effect is achieved,and a more continuous and stable target face is obtained.(3)By identifying and positioning eyes and mouth,the fatigue characteristic parameters were selected and established.Based on the data collected on the actual vehicle,the fatigue detection system is analyzed and designed to achieve the accurate detection of the fatigue state,based on the difference effectiveness of the fatigue characteristic parameters in different fatigue states.The experimental results show that the system has a high recognition accuracy rate,can detect the driver’s fatigue status in real time and accurately.It is of great significance to personal safety protection and traffic safety control.