objektopptak

Objektopptak, often translated as object capture, is the process of creating a digital representation of a real-world object. The result typically includes accurate geometry, surface texture, color, and sometimes material properties, enabling use in computer graphics, virtual and augmented reality, product visualization, archival preservation, and research.

The most common techniques are photogrammetry, structured-light scanning, and laser scanning. Photogrammetry uses many photographs taken around the object and computer vision to reconstruct a 3D mesh and texture. Structured-light scanners project known patterns and measure deformities to recover geometry. Laser scanners emit laser pulses and record return times to generate dense point clouds. For internal parts or microstructures, computed tomography may be used. Some workflows combine techniques to improve accuracy or capture texture.

A typical workflow includes planning, capture, alignment and reconstruction, texturing, and post-processing. Outputs commonly include 3D models in formats such as OBJ, STL, PLY, or FBX; textures as JPEG or PNG; and sometimes point clouds in LAS or LAZ. The data can be integrated into CAD, animation pipelines, or game engines. Proper scale and units must be established, and models may require retopology or decimation to fit real-time applications.

Applications span museums and cultural heritage for preservation, product design and marketing, e-commerce, film and game production, virtual showrooms, and robotics or manufacturing. Challenges include handling reflective or transparent surfaces, occlusions, lighting conditions, object size, motion during capture, and the substantial data size and processing requirements.