People Recognition for Loja ECU911 applying artificial vision techniques

Abstract

This article presents a technological proposal based on artificial vision which aims to search people in an intelligent way by using IP video cameras. Currently, manual searching process is time and resource demanding in contrast to automated searching one, which means that it could be replaced. In order to obtain optimal results, three different techniques of artificial vision were analyzed (Eigenfaces, Fisherfaces, Local Binary Patterns Histograms). The selection process considered factors like lighting changes, image quality and changes in the angle of focus of the camera. Besides, a literature review was conducted to evaluate several points of view regarding artificial vision techniques.

Keywords

• OPENCV
• QT CREATOR
• EIGENFACES
• FISHERFACES
• LBPH
• ICONIX
• FRAME
• artificial vision
• peo- ple recognition
• people recognition