This project focuses on 3D geometric reconstruction from multiple 2D images, aiming to recreate accurate three-dimensional structures using various 2D projections. By leveraging mathematical models and algorithms, the project extracts depth and spatial information from flat images to reconstruct a 3D scene. This technique has applications in fields such as computer vision, robotics, and virtual reality, where understanding the 3D geometry from limited viewpoints is crucial. The project explores the principles of epipolar geometry and triangulation to achieve precise and reliable 3D reconstructions.