3D reconstruction covers methods that compute a three-dimensional model of an object or space from photos, video or sensor data. Typical approaches match several images taken from different angles or evaluate depth data from laser and depth sensors. It is used in surveying, heritage conservation, medicine, robotics and e-commerce – for instance, turning a smartphone scan into a rotatable 3D product model.
Computer vision is a field of AI in which computers interpret images and video. Typical tasks include detecting and locating objects, reading text in images and measuring components. The technology is used, for example, in quality control in manufacturing, in driver assistance systems, in logistics and in medical imaging.
Depth estimation determines from images how far objects or individual pixels are from the camera. This can be done with two cameras (stereo), with depth sensors or, using neural networks, even from a single photo (monocular depth estimation). It is needed in robotics, driver assistance, augmented reality and 3D models of rooms.
Pose estimation is a computer vision technique that determines the posture of people, or the position and orientation of objects, in images or video. Typically a model detects key points such as shoulders, elbows and knees and joins them into a skeleton. It is used, for example, in fitness apps that check exercises, in workplace ergonomics analysis, in sports analytics and in robotics.
