Accuracy Assessment of 3D-Point Clouds with Mobile Mapping System

Abstract

Mobile Mapping System (MMS) is well-known as a modern surveying technology that utilizes to acquire the environmental data of roads in the form of three-dimensional (3D). Previously, receiving the environmental data on the road was difficult and unsafe due to the traffic on the road. Surveying with MMS provides more capability, fast, time-saving, and precision. However, the component of MMS combines with numerous sensors and acquiring data while MMS is moving; the data may contaminate with errors due to several factors. Therefore, this research investigates the accuracy assessment of 3D-point clouds surveyed from MMS by verifying the positioning accuracy using the checking points (CPs) with global navigation satellite system (GNSS)-based positioning through the real-time kinematic (RTK) technique. The examined results of 3D-point clouds before adjusting by ground control points (GCPs) show that the RMSE_r in horizontal is about 0.084 m, and the RMSE in vertical is about 0.123 m. The 3D-RMSE at a 95% confidence level is approximate 0.241 m and falls within Class 3 of NSSDA. And the examined results of 3D-point clouds after adjusting by GCPs indicate that the RMSE_r in horizontal is about 0.076 m and the RMSE in vertical is about 0.046 m. At the same time, the 3D-RMSE at a 95% confidence level is only 0.089 m and falls within Class 2 of NSSDA. The 3D-point clouds have a significant amount of error in vertical, but after adjusting with GCPs, the obtained result shows that the error is significantly decreased.