Advancements in Point-Cloud Processing for Geo-Environmental Modeling: A Case Study of The Dalles Dam Digital Twin Creation

The integration of point-cloud data in geo- and environmental sciences has become increasingly pivotal, with applications ranging from UAVs, spaceborne and airborne lidars to ground-based lidars and stereo-photogrammetric techniques. This session seeks contributions that delve into challenges related to classification, segmentation, and noise removal in the context of point-cloud data, crucial for facilitating change detection studies. Our study focuses on the Navigational Branch of the ERDC Coastal Hydraulics Laboratory tasked with developing a Digital Twin model for a Dam, exemplifying the complexities involved in creating CAD models of terrain and structures.

To address the intricacies of point-cloud data processing, we employed both open-source and proprietary software solutions—Cloud Compare and Autodesk ReCAP— for noise reduction, ensuring the prepared data is seamlessly integrated into CAD modeling software, specifically Inventor. Surface modeling involved the strategic application of planes on cloud points to generate a foundation for sketching and subsequent solid surface extrusion.

Classification of data points was initiated through the implementation of regions in the noise removal software, facilitating the depiction of various areas on the model. Further, color and material assignment in the CAD software enhanced the identification of distinct part areas. Microstation TopoDOT played a pivotal role in creating a detailed terrain model, complete with physical landmarks and water bodies specific to the Dalles dam site.

The resulting models were exported in the desired file format, ensuring compatibility with sponsor requirements. This case study not only showcases the practical challenges encountered in working with point-cloud data but also highlights effective strategies for noise reduction, classification, and model exportation. The presented methodologies contribute to the broader spectrum of geo- and environmental sciences, emphasizing the significance of accurate point-cloud processing for comprehensive modeling endeavors.