The Indoor Environment Reconstruction Based On Stereo Vision

Navigation and mapping services have become very popular in our daily lives.Much of these services are actually available especially for outdoor environments,however applications in GPS-denied environments are being explored where most of the human activities takes place.The aim of this thesis is to reconstruct an indoor environment based on a stereovision system.Hence,to reconstruct a whole map using an unmanned ground vehicle(UGV)we need an accurate position of this robot.In recent days people combine between tracking and mapping with a simultaneous localization and mapping.Visual SLAM systems aim to estimate the motion of a moving camera together with the geometric structure and appearance of the world being observed.Using cameras to provide Visual Odometry(VO)is an effective way to attain such motion estimation.Visual Odometry is an active field of research in computer vision and mobile robotics.In this thesis,ROS based stereo visual SLAM methods and analyzes their feasibility for a mobile robot application in homogeneous indoor environment will be presented.The algorithm utilizes a ZED stereo camera which enables estimation in true scale and easy startup of the system.A distinguishing aspect of the ORB_SLAM2 algorithm is its utilization of a local map consisting of sparse 3D points for tracking and motion estimation.This results in the full history of each feature being used for motion estimation.Furthermore,our algorithm employs Progressive Sample Consensus(PROSAC)instead of RANdom SAmple Consensus(RANSAC)in order to increase robustness against outliers.Extensive evaluations on the challenging KITTI and New College datasets are presented.KITTI dataset was collected by a vehicle driving in the city of Karlsruhe in Germany,and represents one of the most commonly used datasets in evaluating self-driving algorithms.This thesis first studies the performance of ORB_SLAM2 algorithm using KITTI dataset,next it uses a specified camera which is ZED camera in indoor environment.Then,as ZED camera has his own SLAM where is used in the same environment to build a dense map,finally we compare between trajectories obtained by these two SLAMs and the ground truth.