Crowdsourcing pointclouds trends and technologies

This document summarizes crowdsourcing point clouds and related technologies. It discusses crowdsourcing platforms like Mapillary that harness volunteers to collect geospatial data. Structure from motion techniques are explained, which can estimate 3D structure and camera positions from image pairs. The document also introduces CloudSfm and WebODM, open-source tools for reconstructing point clouds in the cloud from internet photos using structure from motion, even without camera or position data. It concludes by considering possibilities for reconstructing historical camera positions through crowdsourcing.