Research On The Detection Of Human Face And Fatigue Based On Intelligence Mimic Algorithm

Global total road traffic accidents happen about 1 billion times a year, specific gravity is about 90% in the total global accidents, and every year at least 500,000 people are killed and 25 million injured. In China every year people dies in the traffic account at least 20% in the total number in the world. Apparently, how to reduce the rate of traffic accidents in China is a great problem to be solved.Research indicates that people take the 80% responsibility in the traffic accident and the fatigue playing the important role. As a result, It’s important to detect the driver fatigue during the driving to prevent the traffic accidents. This is exactly the research of this article.1. Introduce the background, research significance and status of driver fatigue detection. Point out the main method in the paper is detecting the driver’s fatigue degree from the information getting from the facial picture based on computer vision technology. Combine intelligent algorithm with neural network to establish a specific model for detecting fatigue detection.2. According to the degree of fatigue detection,use the RBF (Radial Basis Function) artificial neural network to model and optimized by intelligent algorithms, such as, GA(Genetic Algorithm), PSO(Particle Swarm Optimization Algorithm), BFO(Bacteria Foraging Optimization). Introduce Artificial Neural networks based on Intelligent optimization algorithm and its advantages.3. Discusses the human face region detection and its questions. On the face region detection, Introduces the face and facial area detection algorithm based on Haar-like features. For the light changes causes the reduce of the face and facial region detection accuracy, put forward a light compensation method, improves the detection accuracy and reliability. For the face and facial region tracking, using the TLD (Tracking-Learning-Detection) method to adapt the wide range of head shaking.4. Explain the fatigue information contented in the facial area image, such as, eyes area, mouth area. In the same time, to increase the driver fatigue degree detection accuracy by the single fatigue information, use the information fusion idea, modeling with the RBF artificial neural network. And use the intelligent algorithms for the optimization to enhance the accuracy of fatigue detection.