Building of classified 3D point clouds of outcrop with automatically segmented images

3D point clouds of outcrops are digital representations of rock exposures used for geological surveying. These datasets often have high spatial density, up to a thousand points per square meter. By integrating georeferenced data into the 3D point cloud and applying remote sensing interpretation techniques, geoscientists can extract geological features and build 3D models. However, interpreting 3D point clouds remains labor-intensive, non-reproducible, and prone to human bias. Convolutional neural networks have been applied to segment the images used to build the 3D models, based on a few labeled training and testing subsets, to reduce the amount of human labour. This work used a U-Net encoder-decoder network architecture to segment images of turbidite architectural elements in the La Jardinera Region, Neuquén Basin, Argentina. Approximately 10% of the images were labeled by an expert interpreter, with half used for training and half for testing the model, yielding % overall accuracy of 82%. The model was then retrained on the full labeled set and applied to the remaining unlabeled images. The final segmented outputs were processed through a photogrammetry pipeline to generate classified 3D point clouds, capturing the spatial distribution of architectural elements within the outcrop.