MSV-Based Face Detection in Color Images

This paper presents a MSV-based clustering method for face detection in color images. The method relies on the assumption that faces are well separated from the background by skin color detection. These faces can be located automatically by the minimum scatter volume (MSV) clustering method. The proposed method adopting gradient descent and iteration to search for the optimum solution of cost function which is the determinate of the within-cluster scatter matrix, and the number of clusters is estimated according to the criteria function of clustering effectiveness among numerous clusters. Experiments illustrate the detail of MSV algorithm and also prove the validity of the proposed method.