Author : Deepak Kumar Gupta and Surya Prakash Dwivedi

Detecting with biometric system is a popular and advanced approach in social area. Face recognition method is advanced technique of them to controlling attacks which can be achieved by inserting wrong data using craft faces, programs (Including viruses) and other crafting software thereby gaining ultimate access. An easy task to snippety the face verification system is to use antecedent selected photographs instead of at a time. A simple and furious technique for geometrical feature detection of several human face organs such as eyes and mouth Thus, genuineness verification is required to develop a safe system to protect like such unwanted tactics. According to resultant facts, the images captured from the 2D and genuine faces have much difference in characteristics features like size, shape, gravity and detailednes. Face conversion having printing quality fluctuation which can easily detects via using micro-texture analysis, energy level detected by analyzing the frequency level and posture is carried by gravity features. We are presentencing a vigorous approach to analyzing gravity, frequency analysis and texture calculation and by using frequency descriptor, gravity-center template matching and Local Binary Pattern respectively. This approach generates a special feature space for group unauthorized accessing, detection and face recognition. Experiments on which we have perform, a globally available general database produced better result and we can clearly justify unmasked faces and craft face in 2-D. A specially defined recognition rate of ≃ 89.7% has been achieved for such faces.

Keywords: Frequency Descriptor (FD); Local Binary Pattern (LBP);Edge extraction face feature extraction; Gravity.

Article published in International Journal of Current Engineering and Technology, Vol.7, No.3 (June-2017)