Orbbec has announced that its Gemini 330 series Stereo Vision 3D cameras are now integrated with NVIDIA’s Isaac Perceptor, a workflow designed for autonomous mobile robots (AMRs) leveraging GPU-accelerated Isaac ROS. The announcement was made in Troy, MI, coinciding with NVIDIA’s general availability announcement at COMPUTEX.
The Gemini 330 series cameras are equipped with built-in depth processing capabilities, delivering high-precision Depth+RGB vision suitable for both indoor and outdoor environments. This integration enhances the depth quality and extends the range of sensing capabilities under various lighting conditions, allowing Isaac Perceptor to produce 3D reconstructions and obstacle cost maps in unstructured environments.
The camera models, including the Gemini 335, 335L, 336, and 336L, function in both passive and active laser-illuminated modes to ensure high-quality depth and RGB data even in challenging lighting conditions. These cameras are equipped with Orbbec’s latest depth engine ASIC, which processes depth algorithms internally, thereby offloading this task from the NVIDIA Jetson Orin module-based compute system. Additional features of the cameras include internal IMU and temperature sensors, a working range of 0.2-10 meters, global shutter image sensors, wide field-of-view lenses, high frame rates, low latency, and precise multi-camera synchronization.
To enhance performance in indoor environments, the Gemini 336 and 336L variants are equipped with NIR bandpass filters. These filters reduce the occurrence of “holes” in depth maps caused by glare from reflective surfaces and “ghost” images from repetitive patterns.
Amit Banerjee, Head of Partnerships at Orbbec, emphasized that the integration with NVIDIA Isaac Perceptor provides robot developers with the necessary Depth+RGB vision capabilities to improve the performance and reliability of AMRs across various sectors. The collaboration also includes Orbbec camera simulation models available through the NVIDIA Isaac Sim platform and NVIDIA Metropolis framework, enabling comprehensive utilization of NVIDIA-accelerated technologies from physical runtime models to simulation environments.
Besides AMRs, the Gemini 330 series cameras are suitable for AI vision applications in robot arms, such as bin-picking, palletization, scanning, and sorting, particularly where reducing glare and holes from glossy surfaces is crucial.
