Unmanned Aerial Vehicle (UAV) and ArcGIS for Shipping Container Counting

Abstract

The current approach in counting shipping containers at port terminal in Malaysia is done manually. This process will incur cost, time and labour intensive. Furthermore, this technical issue has led to delay in container operation. This has made the operational efficiency factor to be questioned. Therefore, promoting an automated approached seems appealing.  Previous studies demonstrate the capability of Unmanned Aerial Vehicle (UAV) images for automatic counting of cars and trees. In this study, combinations of aerial images captured using UAV and geo-processing software, ArcGIS, are promoted towards automated approach for counting shipping containers. The overlapping UAV aerial images are post-processed using photogrammetric technique to create Digital Surface Model (DSM) which represents the ground and above surface feature’s elevations. Then, the constructed DSM is filtered to produce the Digital Terrain Model (DTM) that represents the ground surface only. The container’s candidates will then be isolated by subtracting the DTM from DSM to generate normalized DSM (nDSM) that represents heights of container’s stacks. Knowing the standard size and height of one container, the number of container can be extracted. The ModelBuilder tool provided under ArcGIS made automated geo-processing able to be customized. This paper will discuss the steps taken in the automated process and the benefits towards operational efficiency at container terminals.