Skip to Main Content
In order to achieve good tracking results in single camera environment, the human pose of recognition, based on a skeleton and texture model of video algorithms to identify the body is proposed. Single-camera in the background and complexity of the environment, the target audience can only show 2-D information. The algorithm will be the first skeleton to the human body in accordance with the balance weighted tree modeling approach, combined with the skin texture characteristics of the skeleton structure, texture model. Particle filter used for detection and tracking human body. Experiments show that the skeleton model and texture of video algorithms to identify the body with a high recognition rate and robustness, and can do real-time interaction.