IMPLEMENTATION OF FACE DETECTION ON ROS BASED SURVEILLANCE ROBOT FOR INDOOR SECURITY

Indoor security system like CCTV often failed in ensuring room security from theft. Even massive CCTV installation that cost much could be said still ineffective. This failure caused by two major factor : dead angles and no active processing. In this thesis, a mobile surveillance robot with active c...

Full description

Saved in:
Bibliographic Details
Main Author: Hansen
Format: Final Project
Language:Indonesia
Online Access:https://digilib.itb.ac.id/gdl/view/36563
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Institut Teknologi Bandung
Language: Indonesia
id id-itb.:36563
spelling id-itb.:365632019-03-13T14:27:06ZIMPLEMENTATION OF FACE DETECTION ON ROS BASED SURVEILLANCE ROBOT FOR INDOOR SECURITY Hansen Indonesia Final Project surveillance robot, face detection, Viola-Jones , CLAHE, neural network. INSTITUT TEKNOLOGI BANDUNG https://digilib.itb.ac.id/gdl/view/36563 Indoor security system like CCTV often failed in ensuring room security from theft. Even massive CCTV installation that cost much could be said still ineffective. This failure caused by two major factor : dead angles and no active processing. In this thesis, a mobile surveillance robot with active camera processing was designed to navigate autonomously inside any indoor room. The robot built with modified Turtlebot3 Burger robot with additional camera Raspberry Pi Cam, thermal sensor TPA81, and proximity sensor Infrared Sharp. SLAM mapping and autonomous navigation that overcome the dead angle problem was provided by ROS navigation stack. Active face detection was proposed by using custom trained Viola-Jones model, with pre-processing CLAHE for constrast enhancement, finally filtered by post-processing skin detection by using neural network with Hue and Value color input. The result from experiment showed the proposed face detection algorithm by using ROS yielded great accuracy with faster detection time in surveillance robot implementation. text
institution Institut Teknologi Bandung
building Institut Teknologi Bandung Library
continent Asia
country Indonesia
Indonesia
content_provider Institut Teknologi Bandung
collection Digital ITB
language Indonesia
description Indoor security system like CCTV often failed in ensuring room security from theft. Even massive CCTV installation that cost much could be said still ineffective. This failure caused by two major factor : dead angles and no active processing. In this thesis, a mobile surveillance robot with active camera processing was designed to navigate autonomously inside any indoor room. The robot built with modified Turtlebot3 Burger robot with additional camera Raspberry Pi Cam, thermal sensor TPA81, and proximity sensor Infrared Sharp. SLAM mapping and autonomous navigation that overcome the dead angle problem was provided by ROS navigation stack. Active face detection was proposed by using custom trained Viola-Jones model, with pre-processing CLAHE for constrast enhancement, finally filtered by post-processing skin detection by using neural network with Hue and Value color input. The result from experiment showed the proposed face detection algorithm by using ROS yielded great accuracy with faster detection time in surveillance robot implementation.
format Final Project
author Hansen
spellingShingle Hansen
IMPLEMENTATION OF FACE DETECTION ON ROS BASED SURVEILLANCE ROBOT FOR INDOOR SECURITY
author_facet Hansen
author_sort Hansen
title IMPLEMENTATION OF FACE DETECTION ON ROS BASED SURVEILLANCE ROBOT FOR INDOOR SECURITY
title_short IMPLEMENTATION OF FACE DETECTION ON ROS BASED SURVEILLANCE ROBOT FOR INDOOR SECURITY
title_full IMPLEMENTATION OF FACE DETECTION ON ROS BASED SURVEILLANCE ROBOT FOR INDOOR SECURITY
title_fullStr IMPLEMENTATION OF FACE DETECTION ON ROS BASED SURVEILLANCE ROBOT FOR INDOOR SECURITY
title_full_unstemmed IMPLEMENTATION OF FACE DETECTION ON ROS BASED SURVEILLANCE ROBOT FOR INDOOR SECURITY
title_sort implementation of face detection on ros based surveillance robot for indoor security
url https://digilib.itb.ac.id/gdl/view/36563
_version_ 1822924657802608640