Introduction to 3D Vision-Guided Robotics Solution

Mech-Eye Viewer allows users to adjust the parameters of the Mech-Eye industrial 3D camera according to the characteristics of the target object, and obtain high-quality 2D images, depth maps, and point clouds.

Mech-Vision is a state-of-the-art machine vision software. With its fully graphical interface, advanced machine vision applications such as piece picking, high-precision positioning, assembly, industrial inspection/measurement, and automatic path planning can be completed without writing codes.

Based on the image data captured on-site, this software performs a series of vision processing and outputs the vision results (such as the position and orientation of the target object). In addition, based on the vision results, this software can perform collision-free path planning for the robot, and output the planned picking path.

Mech-Viz is a robot path planning software. It uses the information provided by Mech-Vision, including the point clouds and workpiece positions, to intelligently plan the robot path for picking, carrying, palletizing and depalletizing, and other complex application scenarios.

This software allows users to build a workflow for the robot in a visualized manner and provides a 3D simulation function for validation before using the real robot. It has already been adapted to most major robot brands around the world.

Mech-DLK is machine vision deep learning software. With a variety of built-in industry-leading deep learning algorithms, it can solve many problems that traditional machine vision cannot handle, such as highly difficult segmentation, positioning, and classification.

Through intuitive and simple UI interactions, even without programming or specialized deep learning knowledge, users can quickly implement model training and validation with this software.

The communication component for the vision system (runs in the background) enables data exchange among software within the vision system and provides standardized and customized communication with external devices, such as robots, PLCs, and host computers, which makes vision-guided intelligent robotic applications possible.

In this mode, the vision side has control over the robot, i.e., the vision system acts as the master device, while the robot as the slave device. The vision system sends motion or IO commands to the robot, and the robot continuously listens to and executes these commands. The vision system controls the robot to perform tasks based on the planned path, such as loading or depalletizing/palletizing.

When this mode is used, the robot can be controlled by loading a master-control program to the robot or by using the robot’s SDK (Software Development Kit). This mode supports communication between the vision system and the robot.

In this mode, the vision system does not need to take control of the robot. Typically, the robot side acts as the master device, while the vision system is the slave device. The robot side and the vision system use the same standard communication protocol to communicate (such as TCP socket). The robot side sends requests while the vision system processes them and sends responses (poses and labels of the target objects) back. Depending on the request, the vision system returns either the vision result or the planned picking path. The robot makes further decisions or performs the appropriate tasks according to the responses returned from the vision system.

When this mode is used, you should write the robot interface program (for communicating with the vision system) and the robot application program (for receiving the data returned from the vision system and controlling the robot to perform tasks), and load them to the robot. This mode supports communication between the vision system and the robot, the PLC, or the host computer.