Face recognition based on histogram

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

The main scope of this project is to develop an adaptive system along with a bit of intelligence that is capable of identifying the face of the person that is given as input to the system. This objective is achieved by creating a database of different personnel where each person has around 10 samples of images, where each images has different parameters for instances the facial expressions, front view, side views, features that include eye, color. shape, nose, ears etc. These are represented in the form of histogram. Further, using the phase of correlation (POC) technique, according to which the given images which is a 2D images that is represented in time domain is converted to frequency domain and the histogram made for each sample is compared with the input image. During, the execution time the system is given training and later testing of the system is done to find out how far the system is efficient and effective in recognizing the personnel faces.

Tools used:

MATLAB 2006a.

Aim and Objectives:


Face Recognition Based On Histogram.


  • Develop data base of images of different personnel for references.
  • The database consists of a set of sample images of the personnel, where the samples represent each feature of personnel for instance features, color, facial expression etc.
  • The features are represented in form of bar graph for the system to refer.
  • Training of the system is done to make it an adaptive system.
  • After the training is done the system is used for real time purposes.

Literature Review:

A human face is inherently symmetric and we would like to exploit the symmetry in face recognition. As a biometric, face image is the least intrusive, but several challenges are still in process of improving the accuracy of face recognition under illumination changes, variations in pose, occlusions (including self-occlusion), image resolution and other such difficulties (Josh Harguess and J. K. Aggarwal, june 2009). The importance of utilising biometrics is to establish personal authenticity and to detect impostors is growing in the present scenario of global security concern. Development of a biometric system for personal identification, which fulfills the requirements for access control of secured areas and other applications like identity validation for social welfare, crime detection, ATM access, computer security, etc. is felt to be the need of the day. Face recognition has been evolving as a convenient biometric mode for human authentication for more than last two decades (Soumitra Kar, Swati Hiremath, Dilip G. Joshi, Vinod.K.Chadda and 1Apurva Bajpai, august 2007).

Problem Scenario:

The problems that arise during the execution of the system are the size or the pixel definitions of the given 2D input image, as different images have different specifications of pixels its difficult for the system to analysis the specification. The major problem is the method comparison of the images with the database as every image has its own specifications.

Proposed Solutions:

To achieve the optimum output of the system, the first analysis is to confine or fix the parameters of the pixel or size of the given image, which is taken as standard parameters for all the given input images, i.e, around (92 x 112). The second analysis is the comparison of images, which is a complied by using Phase only correlation technique (poc) where the input 2d image that is represented in time domain in x& y axis is converted in frequency domain and during the execution time each input image is correlated with the image in database where the input image is compared each and every sample images present in database, this comparison is easy by converting the image in frequency domain. Using cross- correlation process the two images are compared and accurate results are being displayed.


  • A Multi-Algorithmic Face Recognition System. #Soumitra Kar, Swati Hiremath, Dilip G. Joshi, Vinod.K.Chadda and 1Apurva Bajpai. EISD, BARC, Mumbai-400 085.
  • Robust Motion Estimation for Video Sequences Based on. Phase-Only Correlation, Loy Hui Chien and Takafumi Aoki Graduate School of Information Sciences Tohoku University Aoba-yama 05, Sendai, 980-8579, Japan. email: loy@aoki.ecei.tohoku.ac.jp.
  • IEICE trans.fundamental vol8,march 2004, a finger print matching algorithm using phased only correlation. By loy hui chein and tafafumi aoki.
  • One-Time Key Based Phase Scrambling for Phase-Only Correlation between Visually Protected Images,Izumi Ito (EURASIP Member) and Hitoshi Kiya (EURASIP Member) Graduate School of System Design, Tokyo Metropolitan University, 6-6 Asahigaoka, Hino-shi, Tokyo 191-0065, Japan. Received 24 March 2009; Revised 18 August 2009; Accepted 23 October 2009.