The latest articles on DEV Community by humnaattique4-sys (@humnaattique4sys). https://dev.to/humnaattique4sys https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3999455%2F824065d3-d3e7-4177-9f6e-0ab7aae1dc5b.png https://dev.to/humnaattique4sys en humnaattique4-sys Tue, 23 Jun 2026 21:39:27 +0000 https://dev.to/humnaattique4sys/what-i-learned-about-3d-reconstruction-my-first-week-at-preservemyworld-3ddc https://dev.to/humnaattique4sys/what-i-learned-about-3d-reconstruction-my-first-week-at-preservemyworld-3ddc <p>I just started my internship at PreserveMy.World, a project focused on digitally preserving cultural heritage sites using technology. This is my first week and I spent it setting up my environment and diving into 3D reconstruction research.</p> <h2> What is PreserveMy.World? </h2> <p>PreserveMy.World aims to digitally preserve real places — historical buildings, heritage sites, and landmarks — so they can be experienced even if they're damaged or lost. 3D reconstruction is at the heart of this mission.</p> <h2> The 5 Methods I Researched </h2> <p><strong>1. COLMAP / Structure-from-Motion</strong><br> Takes multiple photos of the same place from different angles and builds a 3D point cloud. Good for reconstructing buildings from tourist photos.</p> <p><strong>2. NeRF (Neural Radiance Fields)</strong><br> Uses a neural network to learn how a scene looks from any angle. Produces photorealistic results but needs a strong GPU.</p> <p><strong>3. Gaussian Splatting</strong><br> A newer, faster alternative to NeRF. Produces high quality real-time 3D scenes with fewer photos.</p> <p><strong>4. Monocular Depth Estimation</strong><br> Estimates depth from just ONE photo. The easiest to get started with — no special equipment needed.</p> <p><strong>5. Multi-View Stereo</strong><br> Uses multiple calibrated photos to build a dense 3D mesh. Good for detailed artifact models.</p> <h2> My Experiment — Monocular Depth Estimation on Google Colab </h2> <p>I decided to start with monocular depth estimation since it only needs one image and runs free on Colab.</p> <p>I used the Depth-Anything-V2 model from HuggingFace to run depth estimation on a synthetic image I created with Python.</p> <p>The result was really cool — the model correctly identified which parts of the image were closer and which were further away, shown through a plasma color map (yellow = close, purple = far).</p> <h2> What Failed </h2> <p>My first two attempts failed because I tried downloading images from Wikipedia URLs which got blocked. I fixed it by generating a synthetic image using Python's PIL library instead.</p> <h2> What I Learned </h2> <ul> <li>Monocular depth estimation is a great entry point into 3D reconstruction</li> <li>You don't need expensive hardware to get started — free Colab works fine</li> <li>For PreserveMy.World, this could be used to get quick depth previews of heritage site photos</li> </ul> <h2> GitHub Repo </h2> <p>All my code, research notes, and experiment outputs are in my repo:<br> <a href="https://github.com/HumnaAttique/PMW-day1" rel="noopener noreferrer">https://github.com/HumnaAttique/PMW-day1</a></p> <h2> Next Steps </h2> <p>I want to try COLMAP next week with real photos, and eventually work toward NeRF or Gaussian Splatting for more realistic heritage site reconstructions.</p> 3d python beginners machinelearning