Since GPS signals are often challenged in indoor or urban environments, many researchers focus on developing simultaneous localization and mapping (SLAM) algorithms using onboard sensors for robots or mobile devices. Nowadays regular cameras are the most available and inexpensive sensor which relies on visual SLAM algorithms. RGB-D cameras become more available but are unreliable in strong sunlight. A lidar is often considered as the most reliable mapping sensor, but it is too power hungry, bulky, and expensive for many mobile robots or devices. A mobile device or a small robot often can only rely on a monocular camera due to power and size constraints.
展开▼
