18-06-2014, 03:05 PM
FACE RECOGNITION USING NEURAL NETWORKS
FACE RECOGNITION USING NEURAL.pptx (Size: 271.54 KB / Downloads: 13)
INTRODUCTION
Facial recognition is a form of computer vision and it automatically identifies a human face from database images.
The face recognition technique consists of three parts-
pre process,
feature extraction and
recognition
RECOGNITION/VERIFICATION MODULES.
Image acquisition/ face detection
Detects face in real time.
The background or the scenes , unrelated to face will be eliminated
Face recognition / verification module
Comprises of pre processing, feature extraction and
Classification sub-modules.
The input is the face image, which is derived from two sources- camera and database
FACE RECOGNITION METHOS
Holistic matching methods
Feature based matching method
Hybrid methods
ALGORITHM FOR FACE RECOGNITION
Measures key points of the face, head angle, skin tone, lighting and creates a template.
Similarity rating
Software allows a user to create their own biometric face identification security for windows
FUTURE WORKS
Algorithms without constrained conditions
Accuracy enhancement.
Unobtrusive audio and video recognition systems.
High recognition rates
APPLICATION
Advanced 3D Facial recognition
Centralized and secure facial identity management solution.
Powerful 2D to 3D transformation of video or photo facial images.
Mobile and video identity resolution surveillance application.
Military and Criminal Investigation
CONCLUSION
we conclude that the computer based face recognition system is very useful for police, industries and for various security regions.
It is less time consuming.
It has numerous applications in areas like surveillance and security control systems.