Automatic Face Recognition via Local Directional Patterns

Automatic facial recognition has many potential applications in different areas of humancomputer interaction. However, they are not yet fully realized due to the lack of an effectivefacial feature descriptor. In this paper, we present a new appearance based feature descriptor,the local directional pattern (LDP), to represent facial geometry and analyze its performance inrecognition. An LDP feature is obtained by computing the edge response values in 8 directions ateach pixel and encoding them into an 8 bit binary number using the relative strength of theseedge responses. The LDP descriptor, a distribution of LDP codes within an image or imagepatch, is used to describe each image. Two well-known machine learning methods, templatematching and support vector machine, are used for classification using the ORL female facialexpression databases. Better classification accuracy shows the superiority of LDP descriptoragainst other appearance-based feature descriptors. Entropy + LDP + SVM is as an improvedalgorithm for facial recognition than previous presented methods that improves recognition rateby features extraction of images. Test results showed that Entropy + LDP + SVM, methodpresented in this paper, is fast and efficient. Innovation proposed in this paper is the use ofentropy operator before applying LDP feature extraction method. The test results showed that theapplication of this method on ORL database images causes 3 percent increases in comparisonwith not using entropy operator.