Multi-Pose Face Kecognition Using Fuzzy Ant Algorithm and Center of Gravity Search

摘要

In this paper, we present the novel technique to solve the recognition errors and minimize memory size of invariant range image multi-pose face recognition. Range image face data (RIFD) was obtained from a laser range finder and was used in the model to generate multi-pose. The fuzzy ant clustering algorithm is used to classify and find the number of clusters for reduced recognition time. RIFD will be transformed by the gradient transformation into significant feature and matching by using Membership Matching Score (MMS) and Center of Gravity (CG) search. The proposed method was tested using facial range images from 130 persons with normal facial expressions. The processing time of the recognition system is better than 3LMS. Moreover, it is 6 times faster without any change of recognition rate. Memory size of this experimental was about 136,890 bytes which show that the memory sizes decrease to extremely small-scale knowledge base.