Leopard Imaging has taken a significant step forward in robotic perception with the launch of its Eagle series stereo cameras, introducing what it calls the world’s first RGB-IR active stereo camera designed for reliable depth sensing in both daytime and nighttime conditions. With the debut of the Eagle 2, the company aims to address one of the most persistent challenges in robotics achieving consistent, high-accuracy vision across changing lighting environments.
At the core of the Eagle series are dual 5.1-megapixel, automotive-grade global shutter RGB-IR sensors. These sensors, combined with active infrared illumination and an integrated dot projector, enable accurate depth sensing not only in bright daylight but also in low-light and even no-light scenarios. As a result, robots equipped with the Eagle cameras can maintain perception accuracy regardless of environmental conditions. In addition, the cameras output synchronized RAW RGB-IR data, allowing developers to perform high-fidelity imaging and precise 3D reconstruction without compromising detail.
AI Authority Trend: Muqing Li Advances Robotic Vision for Autonomous Solar in the AI and Renewable Era
Moreover, Leopard Imaging has integrated NVIDIA Holoscan Sensor Bridge into the Eagle series to ensure efficient data handling. This technology supports long-distance, high-bandwidth data transmission while maintaining low latency, which is critical for real-time robotic decision-making. By enabling synchronized and reliable sensor data flow, the Eagle cameras help robotics systems respond faster and operate more safely in dynamic environments.
At the platform level, the Eagle series benefits from NVIDIA Holoscan, a near real-time AI computing solution built to accelerate edge applications in healthcare, robotics, and industrial automation. Holoscan allows seamless integration of sensors and AI models, delivering near real-time insights and improving operational efficiency. With its optimized sensor interfacing, flexible configuration options, and efficient Ethernet packetization, the embedded Holoscan Sensor Bridge further enhances the Eagle cameras’ suitability for next-generation robotics that demand high perception accuracy and minimal latency.
AI Authority Trend: RoboSense Launches AC1: The All-in-One AI-Powered Robotic Vision Platform
In parallel, the NVIDIA Jetson Thor series provides the compute foundation that elevates the Eagle series even further. Setting a new benchmark for AI robotics performance, Jetson Thor delivers up to 2070 FP4 TFLOPS and supports 128 GB of memory. This substantial leap up to 7.5 times more AI compute than the previous generation allows robots to run advanced generative AI models and multi-sensor perception pipelines simultaneously. Consequently, robots can understand complex 3D environments, react safely, and execute intricate tasks in real time.
Leopard Imaging has designed the Eagle series stereo cameras on the NVIDIA Jetson Thor platform to serve a wide range of demanding applications. These include humanoid robots, physical AI systems, medical and surgical platforms, autonomous mobile robots (AMRs), factory automation, and agricultural robotics industries where robust perception and ultra-low-latency computation are essential.
Looking ahead, Leopard Imaging plans to demonstrate the full capabilities of its imaging portfolio at CES 2026. At the event, the company will showcase multiple NVIDIA-powered imaging solutions, highlighting how its Eagle series cameras are shaping the future of intelligent, perception-driven robotics.
AI Authority Trend: Gemini Robotics 1.5 Marks New Era of Vision-Language-Action in Robotics
To share your insights, please write to us at info@intentamplify.com
