Project Goals and User stories

Product Mission

Certain algorithms and techniques involved in the generation of point clouds can use too many resources or are inefficient depending on various elements.

Our project mission is to evaluate the technologies involved in the generation of point clouds and optimize a factor in its pipeline

Our main interest is to improve the performance of reconstruction using deep learning

Our focus at large is to allow field scanning work in heterogeneous environments to either be sped up or for computation on the edge to be adequate

Users

Users who are well versed in the field of photogrammetry

• • Researchers, Students, Professors, Universities
• • Start-up companies Users who are not well versed in photogrammetry
• • Investors
• • Manufacturers
• • Restoration enthusiasts
• • Architects
• • Land Surveyors

User Stories

• • As a researcher, I would like to utilize this software to determine how well the optimization of the photogrammetry pipeline worked
• • As an investor, I would like to understand the optimization of photogrammetry clearly without much prior knowledge
• • As a manufacturer, I would like to utilize the output of this software directly in 3D printing
• • As an architect, I would like to utilize this software without the need for a high-end computer to achieve building scans

Old MVP

Our MVP is to improve the performance of a factor in the photogrammetry pipeline using machine learning

• • Optimistically, we would like to create this machine learning method in Pytorch as its a very mature machine-learning framework
• • One of our group members also owns a powerful Nvidia graphics card so local development would be a lot easier on Pytorch
• • As we explore the world of photogrammetry we will update the MVP with our preferred factor in the pipeline to optimize. For now, we are interested in mesh reconstruction

New MVP

Densify the point cloud that could be fed to an AI or user

• Users: any company that uses point clouds in open source

MVP User Stories

• As a researcher, I would like to have all the necessary information to install the script in an easy to access repository to evaluate our group testing methods for optimization of the photogrammetry algorithm.
• As an employee in a startup company, I would like to run this software (preferably as a .exe) and have the script explain how well the optimization of a factor in the photogrammetry pipeline worked
• As a company, I would like to densify the point cloud that could be into our open source software