An interactive robotic arm with learning function

We have developed a robotic arm for intelligent and automatic gripping of objects. Its working principle is to manually input instructions, and then the system processes the instructions to remove invalid information, and then divides the effective information into two parts: the action and the target object. The action instruction The target command is transmitted to the motion system, and the target command is transmitted to the vision system. The system first pre-processes the collected images, and then performs feature extraction to find the target object. When the target object is confirmed, the vision system uses the principle of binocular ranging to measure out The coordinate parameters of the camera and the target object, the system processes the parameters, obtains the motion parameters of the steering gear and the motor, and then transmits them to the steering gear and the motor. After the steering gear and the motor start, the mechanical arm moves. When the movement stops, the visual The system measures the distance again, and the system judges the measured parameters. If the parameters meet the gripping conditions, the grippers on the robotic arm perform the gripping operation. If the parameters do not meet the gripping conditions, repeat the distance measurement, movement, and measurement. distance until the coordinate parameters of the object and the manipulator meet the gripping requirements.