The latest articles on DEV Community by Ayesha Noor (@ayesha_noor_fd15fc0aed408). https://dev.to/ayesha_noor_fd15fc0aed408 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%2F4000493%2Fbc9d263d-6dd9-49d4-9377-c8f578a09141.jpg https://dev.to/ayesha_noor_fd15fc0aed408 en Ayesha Noor Wed, 24 Jun 2026 11:34:59 +0000 https://dev.to/ayesha_noor_fd15fc0aed408/how-3d-reconstruction-can-preserve-cultural-heritage-my-week-1-with-preservemyworld-42og https://dev.to/ayesha_noor_fd15fc0aed408/how-3d-reconstruction-can-preserve-cultural-heritage-my-week-1-with-preservemyworld-42og <p>As part of my Week 1 internship with PreserveMy.World, I dove into the world of <br> 3D reconstruction and how it can be used to digitally preserve heritage sites.</p> <h2> What is 3D Reconstruction? </h2> <p>3D reconstruction is the process of capturing the shape and appearance of real <br> objects or environments and converting them into digital 3D models. For heritage <br> preservation, this means we can create permanent digital records of sites that may <br> be at risk of damage, decay, or destruction.</p> <h2> Methods I Researched </h2> <p><strong>COLMAP (Structure from Motion)</strong></p> <ul> <li>Input: Multiple photographs from different angles</li> <li>Output: 3D point cloud and mesh</li> <li>Difficulty: Medium — good beginner starting point</li> </ul> <p><strong>NeRF (Neural Radiance Fields)</strong></p> <ul> <li>Input: Photos with camera position data</li> <li>Output: Photorealistic 3D scene you can navigate</li> <li>Difficulty: Hard — requires GPU and ML knowledge</li> </ul> <p><strong>Gaussian Splatting</strong></p> <ul> <li>Input: Photos</li> <li>Output: Real-time renderable 3D scene</li> <li>Difficulty: Hard — very new technique (2023)</li> </ul> <p><strong>Monocular Depth Estimation</strong></p> <ul> <li>Input: A single image</li> <li>Output: Depth map showing distances</li> <li>Difficulty: Easy — great for beginners</li> </ul> <h2> Why This Matters for PreserveMy.World </h2> <p>PreserveMy.World aims to use AI to document and preserve heritage sites. <br> 3D reconstruction gives us the ability to let people virtually visit places <br> that may no longer exist. Imagine being able to walk through an ancient site <br> in VR, reconstructed from photographs taken before it was damaged.</p> <h2> What I Plan to Try Next </h2> <p>This week I plan to run a COLMAP demo or find a beginner-friendly Colab notebook <br> to test structure-from-motion on sample images. I'll document the results and <br> share outputs in my PMW-day1 GitHub repo.</p> <h2> My Setup </h2> <ul> <li>GitHub repo: <a href="https://github.com/i232103-dot/PMW-day1" rel="noopener noreferrer">https://github.com/i232103-dot/PMW-day1</a> </li> <li>Track: AI &amp; Machine Learning — AI-Based 3D Scene Reconstruction</li> <li>Program: PreserveMy.World Internship, Week 1</li> </ul> ai beginners deeplearning devjournal