Indoor Localization Algorithm Based On Visual SLAM

Simultaneous Localization and Mapping(SLAM)is a key technology in the field of robot,which has a wide range of applications.Traditional visual SLAM is mostly based on point features,and it is relatively mature in extraction and description algorithms.However,point feature are more dependent on the environment,and are not performing well in textureless scenes and etc.The man-made buildings exhibit strong structural regularity and in most cases can be abstracted as blocks that are stacked together with three dominant directions,which is known as Manhattan-world assumption.The lines which are aligned with the dominant directions are called structure lines.Unlike other line features,the building structure lines encode the global orientation information that constrains the heading of the camera over time,eliminating the orientation drift caused by the accumulation of angular errors and consequently reducing the position drift.In this thesis,we proposes a visual SLAM method based on point and structural line features extracted from the image,which can be applied to locate and estimate the robot in unknown indoor environments.The main research work includes:1.An improved extraction and matching algorithm of structural lines.Firstly,straight line detection algorithm and vanishing point extraction algorithm are used to complete the extraction of structural line features.Then use line segment alignment to match the line segments in the inter-frame image,2.In the back-end of the proposed visual SLAM,we construct the graph model using point and line feature.And we employ the orthonormal representation to parameterize lines and analytically compute the corresponding Jacobians.3.We design and implement a complete monocular visual SLAM system using point and structural line features,which includes frame tracking,local mapping,bundle adjustment of both line feature and point feture.Extensive experimental results are present to validate its performance.