Common types of cameras used in general humanoid robots

Binocular cameras mimic human binocular vision, capturing the object from different angles using two cameras, and providing depth perception capability by using the principle of parallax. This allows the robot to sense the three-dimensional spatial position and of objects, achieving a three-dimensional perception of the environment. When the robot grasps items, binocular cameras can accurately measure the position and posture of the items, the accuracy and success rate of grasping.

Depth cameras can directly obtain the depth information of objects. Common depth camera technologies include structured light and time-of (ToF). Structured light depth cameras project specific light patterns onto the surface of objects and calculate the depth of objects based on the deformation of the patterns; ToF cameras, on the other hand, calculate the distance of objects by measuring the time it takes for light to travel from emission to return. Depth cameras provide robots with more comprehensive environmental perception capability and play an important role in tasks such as robot navigation, obstacle avoidance, and object recognition.

Infrared cameras can capture infrared radiation emitted by objects and are not limited by light conditions, allowing them to operate normally in low-light or dark environments. In fields such as security monitoring and industrial, infrared cameras help general humanoid robots obtain thermal images of target objects, enabling temperature monitoring, thermal target tracking, and visual perception in dark environments.

Ordinary RGB cameras capture image information through three color channels: red (R), green (G), and blue (B), and can present rich colors and delicate texture. This type of camera is suitable for scenarios where high requirements are placed on the appearance and color recognition of objects, such as face recognition, object classification, and image, providing robots with basic visual perception of the surrounding environment.