The Design And Research Of Face Recognition Algorithm On Vehicle

In a large number of biometric identification technology, face recognition technology has the advantage of uniqueness, stability, reliability and non-contact. We can use face recognition technology to the anti-theft on vehicle. It will greatly enhance the security of vehicle. In this paper it improved face detection algorithm based on skin color and proposed the face recognition algorithm based on Spiking Cortical Model (SCM) and Composite Correlation Filter. The algorithms have achieved good results.In this paper it improved face detection algorithm based on skin color. In order to adapt the car environment, it made color region classed by Gaussian model. After the operation which is binaryzation and expansion-erosion-expansion, it had removed the interference region and then determined the region of face according to the ratio of length and width. Calculate the degree between the line which determined by barycenter of two eyes and horizontal line. Through the degree it can judge whether the face is parallel. If the face is not parallel rotate the region of face. At last, normalize the image of face.In this paper, it proposed a face recognition algorithm based on Spiking Cortical Model (SCM) neural networks and Composite Correlation Filter. SCM segments face image though multiple iterations. It extracted face feature though iteration images. It processed the feature vector though Composite Correlation Filter to match with the vectors in the feature lib. It determined the recognition result according the PSR value of matching.Apply the algorithms in experiments of vehicle. The result showed the algorithms were effective.