An Improved Linear Blending Skinning Algorithm Based on Human Joint Position Relation and Improved Weight Calculation Method

Abstract

Based on real-time requirements and joint position information in motion, this paper improves LBS algorithm because of its highest computational efficiency. The weight calculation method in traditional LBS algorithm is improved into a pair of values—The ratio of The length of the vertex's projected position on the bone to the length of the bone and Cosine of the Angle between the attachment of the vertex to the parent joint and the bone containing the parent joint--to Overcome the wrap candy paper and articulatory place collapse. After extracting the skeleton structure of the model and calculating the weight, by monitoring the position change relation of the local key joints in the motion sequence, the traditional LBS algorithm and the improved LBS algorithm can be dynamically switched to reduce the computation as much as possible and improve the algorithm efficiency. Experimental results show that the improved algorithm proposed in this paper overcomes the defects of LBS algorithm effectively, and the method is simple and effective.