Drone related Articles and tips

Steps required to create a 3d model using photogrammetry

Steps required to create a 3d model using photogrammetry

Creating a 3D model from a set of images involves a process called "photogrammetry." Photogrammetry is a technique that uses photographs taken from different angles to reconstruct a 3D representation of an object or scene. Here's a general overview of the steps involved in creating a 3D model through photogrammetry:

  1. Image Capture: The first step is to capture a series of high-quality images of the object or scene you want to model. These images should be taken from various angles, providing good coverage of all sides and details of the subject.

  2. Image Preprocessing: Before creating the 3D model, it's essential to prepare the images for photogrammetry. This may involve resizing, cropping, and correcting any distortions or color imbalances in the pictures. Software tools designed for photogrammetry often include features to automate this preprocessing step.

  3. Feature Extraction: In this step, the photogrammetry software identifies and matches distinctive features (known as keypoints) in the images. These features are points or patterns that can be easily identified in multiple images, providing a way to correlate different viewpoints of the same object.

  4. Correspondence and Triangulation: Once the software identifies matching features across images, it calculates the relative positions and orientations of the camera viewpoints. This process, known as triangulation, helps determine the 3D coordinates of the features in the real world.

  5. Point Cloud Generation: By combining the 3D coordinates from multiple images, the software creates a point cloud. A point cloud is a collection of 3D points representing the surface of the object or scene. It provides a sparse representation of the 3D model.

  6. Surface Reconstruction: To create a solid 3D model, the point cloud is processed further to generate a mesh representing the surface of the object. Meshes are composed of connected triangles that form the outer shape of the model.

  7. Texture Mapping (optional): If the original images contain color information, the photogrammetry software can project the images onto the 3D mesh, adding realistic textures to the model.

  8. Post-processing: After generating the 3D model, some post-processing may be necessary to clean up artifacts, fill holes, and refine the overall quality of the model. This step often involves using 3D modeling software to make adjustments and optimize the model.

  9. Exporting and Use: Once the 3D model is complete, it can be exported in various formats suitable for different applications, such as 3D printing, computer graphics, virtual reality, or architectural visualization.

It's worth noting that the accuracy and quality of the final 3D model depend on several factors, including the number and quality of images, the resolution of the camera used, the lighting conditions during image capture, and the performance of the photogrammetry software being employed. More advanced photogrammetry techniques and hardware, such as using structured light scanners or laser scanners, can enhance the accuracy and level of detail in the resulting 3D model.

Drone Media Scotland is not only experienced in Photogrammetry, we are one of the few to have actually studied the subject.