A key challenge in computer vision is to reconstruct the three-dimensional geometry of an item or scene from two dimensional pictures taken with various sensors. The geometric structure of real-world objects may be recovered using a variety of approaches that are used in gaming, medical imaging, mechanical component inspection, object identification, and other fields. The precision of the reconstruction, the complexity of the calculation, and the speed all affect how well 3D reconstruction algorithms function. Passive and active 3D reconstruction techniques are examined in the quest to create an effective approach, and a few well-known techniques are selected based on the knowledge found in the literature at the time. To enhance 3D reconstruction performance, these techniques are updated. The goal of this project is to provide approaches for reconstructing the 3D structure of real-world objects or scenes from photographs taken with various sensors and to assess how well each approach performs in terms of accuracy, complexity, and speed. It is challenging to get precise depth information from a single picture, hence two or more photos must be used to complete 3D reconstruction. In the suggested study, stereo vision is used to execute a passive approach of 3D reconstruction. To determine the depth of a scene or an item, two photos are taken. Scale Invariant Feature Transform (SIFT) is used to extract and match key points. Since SIFT produces more key points than other feature extraction techniques, it is used because it produces more features overall. The camera calibration toolkit is used to calibrate cameras. Epipolar rectification is carried out after calculating the basic matrix. The region-based stereo method with global error energy reduction is used to create the disparity map. The disparity map is then used to create the depth map. The depth information is not retrieved in certain locations when the findings are compared to ground truth values. The 3D reconstruction is carried out using numerous photos to get over this limitation. Newline