From OrbbecReviewed by Lexie CornerJun 5 2024
Orbbec, an industry leader dedicated to 3D vision systems, announced its Gemini 330 series Stereo Vision 3D cameras are now integrated with NVIDIA Isaac Perceptor, a reference workflow for autonomous mobile robots (AMRs) built on GPU-accelerated Isaac ROS.
These cameras enhance depth quality and provide longer-range sensing in varied lighting conditions, which lets Isaac Perceptor – whose general availability was announced by NVIDIA today at COMPUTEX – output 3D reconstruction and obstacle cost maps of any unstructured environment.
The Gemini 335/335L/336/336L cameras operate in both passive and active laser-illuminated modes to ensure high-quality depth and RGB data output even in challenging lighting conditions. The depth algorithms are processed in the camera by Orbbec’s latest depth engine ASIC and thus eliminates the burden on the NVIDIA Jetson Orin module-based compute for such operations. The cameras include internal IMU and temperature sensors and have 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.
Orbbec also announces the Gemini 336/336L variants for improved performance in indoor environments by adding NIR bandpass filters. This reduces the potential of “holes” in a depth map due to glare from shiny floors and other reflective surfaces and “ghost” images from repetitive patterns in the environment.
“The integration of Orbbec’s cameras with NVIDIA Isaac Perceptor will provide robot developers the precise Depth+RGB vision capabilities necessary to enhance the performance and reliability of AMRs across various sectors,” said Amit Banerjee, Head of Partnerships at Orbbec. “This collaboration extends beyond physical applications, with Orbbec camera simulation models also available through the NVIDIA Isaac Sim platform and NVIDIA Metropolis framework to ensure customers can fully utilize the capabilities of NVIDIA-accelerated technologies from physical run-time models to physically-based simulation environments.”
In addition to AMRs, the Gemini 330 series cameras are well suited for robot arm applications that utilize AI vision for bin-picking, palletization, scanning and sorting applications, especially where reduction in glare and resulting holes from glossy surfaces are important.