Abstract: To deal with the problem of rapid eye localization in low quality images, a rapid eye localization algorithm combining separability filter and local binary pattern (LBP) is proposed in this paper.This algorithm can calculate the multi-scale local binary patterns (LBPs) value effectively, and combine the separability filter and LBP with the idea of cascade to locate the eyes rapidly and accurately.The simulation results in the BioID face database show that the average time consume is decreased by nearly 40% compared with the common LBP algorithm while the accuracy is guaranteed at two scales.