Robot Terrain Inclination Model Extracted From Laser Scanner Dada For Outdoor Environment

In the last decades, laser scanning systems are used widely to extract theobject information in the open areas.The laser scanner can be used to do localizationor path planning in robot reseach field.To operate in the outdoor environment, robotsneed the appropriate data structures to support.But it is really hard to get the real andefficient information in the outdoor environment for it’s complicated surroudings.Although three-dimensional laser point clouds can represent the environment withhigh accuracy, the high computational demands prevent them from being directlyapplied in the large-scale outdoor environment. So a new accurate and efficientterrian model proposed in this paper has a profound significance in the outdoorenvironment.In this thesis, only terrain points from laser data will be used to extract a newRobot-Terrain Inclination (RTI) model. RTI model is defined here to describe therelationship between the robot attitude (roll/pitch/yaw) and the position of the robotwhere the roll and pitch of the robot are varied according to the inclination changesof the natural terrain the robot is operating on, and the yaw is determined once thepath is given.In order to extract this model, the segmentation of the terrain pointsand the non-terrain points is proposed.Experiments were carried out to demonstrate the accuracy of RTI model.Comparing the results between the estimated RTI model and the reference model, theproposed method effectively estimates RTI model with reasonable precision. And heproposed method can be used to build an effective RTI model for the inclinedoutdoor terrains. In the near future, the proposed technique can be used to evaluatethe terrain traversability or assist localization and mapping for outdoor mobile robot.