Face recognition application based on Gabor filter Zernike Moment with Matlab Source Code

Download Free projects with Source Code


In modern world, everything becomes computerized. The technology growth demands for accurate and quick user identification. Access key options for anything needs security clearance. Using proper access codes user verification is needed. Even in modern technologies, the secured access keys like pin numbers, passwords can be hacked easily. This drawback can be overcome by using the face recognition technology. Face recognition can immensely help the owner despite in identical twins case. Facial recognition is a part of the biometrics. Biometric is a unique technology which automatically verifies the user details based on their specific identity. Some of the biometric types are, fingerprint recognition, face recognition, voice recognition, hand recognition, iris recognition. The biometrics does both the behavioral characteristics and physical characteristics. The behavioral biometrics is based on the data derived from an action, whereas physiological is based on the data derived from human body part measurements. Facial recognition does not need any interaction with any user; it works completely automatic once the data entered in storage. It allows highly accurate verification rates. The facial recognition includes two processes: verification and identification. 


Image acquisition: This module includes technology that acquires faces from any static camera that generates needed quality and resolution. High quality imaging is necessary for face identification and recognition. This acquisition process defines the refined facial characteristics in future authentication process.

Image processing: In this module, Images been cropped to avoid background remains, color images are converted to black and white to facilitate comparison in grayscale characteristics. First, face should be detected and then it must be localized and normalized for face sample alignment on one template which helps easy identification.

Unique characteristic spotting: In this module, all facial recognition attempts must match the sample facial features similar to people identifying each other. This feature often used in face recognition system to change some significant features like nose shape, mouth sides, eye’s upper ridges, cheek bone areas, growing and shaving facial hairs, with or by removing eyeglasses, changes in make-ups  etc.,

Pattern creation: In this module, the patterns are normally created by processing multiple facial images. These patterns can vary in size. The larger patterns like 3k template are linked with behavioral biometrics.

Pattern equivalent: In this module, this compares the new patterns against stored patterns. A series of images may occur at a time, but the system processes in 1 or 2 seconds. 


Get Code