Similar to the application of 3D perception in the field of smartphones, the current 3D perception technology of robot vacuum on the market also includes structured light, stereo vision (Stereoscopy) and 3D ToF camera. These three solutions are indeed the mainstream of the current optical ranging solutions.
Among them, stereo vision mostly adopts dual-camera or multi-camera schemes, and distance detection is carried out by triangulation method-general stereo vision is a passive optical ranging scheme, because this kind of scheme does not actively emit light, but uses ordinary Camera. Therefore, the application of stereo vision is limited by the lighting environment, and its perception ability will be greatly reduced in the dark environment. As a result, the robot vacuum cannot work efficiently at night or in low-light environments.
Like ToF, structured light also belongs to an active optical ranging scheme, which includes two parts: the transmitting end and the receiving end. However, the technical requirements for the transmitting end of structured light are obviously higher. It needs to emit a light spot containing a certain pattern, and then the receiving end captures the light spot projected into the scene, and calculates the distance based on the deformation of the light spot. The characteristic of this kind of scheme is generally higher precision, but the cost is higher, and the measurement distance is limited by the distance from the receiving end to the transmitting end.
3D ToF camera avoids the shortcomings of the above two mainstream solutions and is a suitable technology. In the previous article, ToF single-point optical ranging was mentioned. 3D ToF camera refers to extending a single point to multiple points to realize the transition to true 3D perception.
In the specific application of robot vacuum, 3D ToF camera has inherent applicability. For example, within the same module volume, its measurement distance is larger than structured light; and it can work in strong light or dark light environment, which is obviously suitable for the environment. The adaptability is higher than stereo vision. Therefore, 3D ToF camera is the best choice for the 3D perception ability of the robot vacuum.
