Using LBP histogram for face recognition on Android platform

In this paper, a new face recognition method based on local binary pattern(LBP) histogram is presented. First, the grey face image is divided into several parts, and then the LBP histogram of each part is calculated and concatenated together. Second, extract other concatenated histograms using different LBP operators and image blocking patterns. All the histograms above are concatenated into the ultimate vector. The proposed approach was tested using Yale face database on Android platform. Experimental results demonstrate that the combined LBP histogram feature has an encouraging performance for face recognition.