Unleashing the Power of Robot Arms and Intel RealSense 3D Cameras

Intel® RealSense™ Depth Cameras are perfect for many robotic applications, such as pathfinding and navigation, human safety, and picking/handling objects.

We love finding interesting Intel RealSense technology use cases in our developer community. One such example is Elephant Robotics and their new educational robotic arm kit, the AI Kit 3D. Empower learning and explore the future with their user-friendly Robotic Arm and Intel RealSense Depth Camera Kit, which masters 3D camera technology.

Their AI Kit has five built-in main visual recognition algorithms powered by Intel RealSense technology. Users can use different-colored wooden blocks and different-shaped cards for the recognition function. Also, users can use four types of QR CODE recognition to learn the relationship between 2D and 3D. Feature point recognition helps users understand image segmentation and features and learn the YOLOv5 algorithm quickly.

The image below shows that the Intel RealSense Depth Camera D455 has four lenses, including an infrared laser projector, two infrared cameras, and a color camera.

These lenses serve specific purposes:

• Infrared Laser Projector: The camera can project an infrared light point grid into the scene, and the infrared cameras capture these points. Because the positions of the projector and cameras are fixed, calculating the offset of the light points in the cameras can infer the distance of each point from the camera, thus obtaining the depth information of the scene.
• Infrared Cameras: Infrared cameras can capture the infrared spectrum, which is part of the electromagnetic spectrum with wavelengths longer than visible light but shorter than microwaves. The main function of infrared cameras is to image in conditions without visible light illumination, as many objects emit, reflect, or transmit infrared light.
• Color Camera: This camera is typically used to capture the standard visual information of the scene, while other cameras capture additional information, such as depth information or images in low-light conditions. This data can be combined with the images captured by the color camera to provide richer visual data, supporting advanced functions like facial recognition, augmented reality, or 3D modeling.

Combining the functionalities of these built-in camera lenses allows for obtaining three-dimensional information about an object. This technology can be used in various applications, including facial recognition, gesture recognition, object recognition, and measuring object depth.

Elephant Robotics’ Artificial Intelligence 3D Kit is an entry-level package for robotic arm applications in artificial intelligence and machine vision. The kit utilizes four recognition algorithms: color recognition, shape recognition, YOLOv8, etc. It is compatible with a visualized operating interface and addresses the limitations of 2D cameras requiring marker positioning using a 3D camera. The open-source code is based on the Python platform, allowing for mechanical arm control through development software. More importantly, it offers a unique opportunity for users to delve into and master 3D camera algorithms.

As technology advances, the use of 3D cameras is rapidly expanding into multiple fields, including—but not limited to—manufacturing, security, entertainment, and healthcare. Kits like this one can help users grasp this technological trend, laying a solid foundation for their future learning and careers.

Check out their Hackster article and GitHub repository to learn more about this excellent product.

If you are a developer building novel technologies with Intel RealSense depth cameras, we would love to hear from you! Send us an email at .