A Method for Facial Recognition Based On Local Features
K. Srinivasa Reddy, V. Venkata Krishna, V. Vijaya Kumar · International journal of mathematics and computation · 2015
Recently many researchers working in the area of image processing shown tremendous interest in deriving precise face recognition methods due to its challenging nature and diverse set of applications. The present paper derived a face recognition model by taking into account the local features derived based on Local Binary Patterns (LBP). The local features of the proposed method are derived based on the combination from a subset derived from Uniform Local Binary Pattern (ULBP) and Non Uniform Local Binary Pattern (NULBP). To overcome the disadvantage of dimensionality problem, many researchers considered only ULBPs for their applications and this has made a huge number of NULBPs as miscellaneous and by which overall performance is degraded. To address this problem the present paper derived a subset called Extended-Significant Non Uniform Local Binary Pattern (E-SNULBP) from the NULBP set. Based on this E-SNULBP and its various combinations, the present paper derived a face recognition algorithm. The proposed method is implemented on Indian, American Telephone and Telegraph Company (AT and T) Olivetti Research Laboratory (ORL) and Yale databases and experimental results indicate a high facial recognition rate.