Development of a Low-Cost MEMS INS/GNSS Integrated System for LIDAR Mobile Mapping Applications

Mobile mapping systems using LiDAR technology are becoming a popular choice to collect 3D data and determine surface elevation models. Mobile mapping systems integrate several advanced surveying technologies. The basic components of these systems include a direct georeferencing component along with a laser scanner (LiDAR) and/or a digital imager. Direct georeferencing is the determination of position and attitude parameters for a mobile mapping system. The most common technologies used for this purpose today are Inertial Measurement Unit (IMU) and Global Navigation Satellite System (GNSS) integrated systems. However, only costly and expensive navigation and tactical grades IMUs have been implemented for direct georeferencing of the LiDAR mobile mapping data. Most recently, the production of low-cost industrial grade IMUs using Micro-Electro-Mechanical System (MEMS) technologies are evolved rapidly to improve the IMU accuracy and to achieve comparable accuracy to the tactical grade IMUs. This research will investigate the development of accurate low-cost MEMS based inertial navigation system (INS)/GNSS georeferencing system for LiDAR mobile mapping applications for 3D data collection and surface elevation models production. The challenge of using the MEMS IMU is that it is contaminated with high level of noise and high level of bias stability error. To overcome this issue, a new denoising, filtering and integration methods will be developed in this research project using hybrid wavelet neural network (WNN) and principal component analysis (PCA) methods to achieve accurate MEMS based INS/GNSS navigation solution for mobile mapping applications. Moreover, the final accuracy of MEMS based INS/GNSS navigation solution will be compared with the American Society of Photogrammetry and Remote Sensing (ASPRS) horizontal and vertical accuracy standards for geospatial data production.