ETD

Digital archive of theses discussed at the University of Pisa

 

Thesis etd-09152016-085435


Thesis type
Tesi di laurea magistrale
Author
JOAO, VALDO ANTONIO
URN
etd-09152016-085435
Thesis title
Parallelizing Image Processing Algorithms For Face Recognition on Multicore Platforms
Department
INFORMATICA
Course of study
INFORMATICA E NETWORKING
Supervisors
relatore Prof. Buttazzo, Giorgio C.
correlatore Prof. Danelutto, Marco
controrelatore Prof. Cisternino, Antonio
Keywords
  • Viola-Jones algorithm
  • parallelization
  • Multi-core Platforms
  • Image processing
  • Face Recognition
Graduation session start date
07/10/2016
Availability
Withheld
Release date
07/10/2086
Summary
Humans are able to easily locate faces in its environment despite difficult conditions such as occlusions of parts of a face and bad lightening. Many studies have been conducted to try to replicate this process automatically, using machines.
In this paper we first analyze the existing Face Detection algorithms that are able to provide real-time processing. A detailed analysis of the Viola-Jones algorithm is made. Several techniques are proposed in order to overcome the Viola-Jones algorithm limitations.
An optimization to the current OpenCV implementation of the Viola-Jones algorithm is proposed along with additional parallelization techniques in order to provide better performance on Multi-core Platforms.
File