Human Recognition at a Distance in Video by Integrating Face Profile and Gait

摘要

Human recognition from arbitrary views is an important task for many applications, such as visual surveillance, covert security and access control. It has been found to be very difficult in reality, especially when a person is walking at a distance in read-world outdoor conditions. For optimal performance, the system should use as much information as possible from the observations. In this paper, we propose an innovative system, which combines cues of face profile and gait silhouette from the single camera video sequences. For optimal face profile recognition, we first reconstruct a high-resolution face profile image from several adjacent low-resolution video frames. Then we use a curvature-based matching method for recognition. For gait, we use Gait Energy Image (GEI) to characterize human walking properties. Recognition is carried out based on the direct GEI matching. Several schemes are considered for fusion of face profile and gait. A number of dynamic video sequences are tested to evaluate the performance of our system. Experiment results are compared and discussed.