Accuracy Assessment of Airborne Mapping Based on Variation of Number and Distribution of Ground Control Points

Abstract

Orthomosaic is a map derived from numbers of overlapped aerial images that are stitched together using photogrammetry software. Orthomosaic can be used as an analysing tool to extract topographic features and monitoring applications. The accuracy of any photogrammetry product is important and can be dictated by various reasons. One of the prominent factors is the utilization of Ground Control Point (GCP). GCPs is known as sets of points in the study area that were established to provide aerial images with known coordinates through georeferencing process. The optimum GCP distribution is at the edge of the study area. However, it is sometimes impossible to access several areas of the mapping area due to unexpected prohibitions. The only option is to establish the point at the area where they can access it. Different numbers and distribution of GCP will contribute to the changes in accuracy. For this study, twelve points were established using GNSS observation. These points are utilized as either GCP or Check Point (CP). The number of GCP for each processing differs from three GCPs to six GCPs. Each number of GCP were distributed either at the edge or centre of the area. Edge distribution is supported with several GCPs at the centre. This study utilized a multi-rotor drone with an integrated camera for data acquisition. The method of analysis for this study is by comparing the coordinates of CP and computing the planimetric and vertical RMSE from each orthomosaic. This study found that the best GCP configuration with optimum accuracy is six GCPs with edge distribution with the RMSE of 2.9 cm and 4.4 cm for planimetric and vertical accuracy, respectively. This study helps to plan the task to establish the GCPs thoroughly to ensure great accuracy can be achieved regardless of the circumstances.