To convert point cloud to realistic 3D models I have developed an algorithm to get an overview of wireframe for that point cloud. My work mainly lies on extracting such points from point cloud which would help us in creating a wireframe for our 3D model. I have developed two methods for edge detection and corner detection.
After conducting several experiments using various machine learning algorithms, it was concluded that DBSCAN outperformed all other algorithms.
As DBSCAN performed well in edge detection, the experiment for corner detection was conducted solely on it.