OMNIVISION, a leading global developer of semiconductor solutions, including advanced digital imaging, analog, and touch & display technology, announced that its total camera solution, comprising the OG02B10 color global shutter (GS) image sensor and OAX4000 ASIC image signal processor (ISP), is now verified and available with the NVIDIA Holoscan sensor processing platform and the NVIDIA Jetson platform for edge AI and robotics. A premier demonstration of the combined solution will take place this week at the VISION 2024 show in the OMNIVISION booth (Hall 10, Booth 10G82), October 8-10, 2024, at Messe Stuttgart in Germany.
Humanoid robots, which resemble and act like humans, represent a growing sector that was valued at $1.8 billion in 2023 and is predicted to increase to more than $13 billion by 2028, according to research firm MarketsandMarkets1. The robots are engineered to imitate human expressions, interactions and movements; to do this, they require an array of cameras, sensors, and artificial intelligence (AI) and machine learning (ML) technologies.
NVIDIA Holoscan is a full-stack AI sensor processing platform for low-latency sensor and network connectivity that includes optimized libraries for data processing, AI and core microservices to run streaming, imaging and other applications, from embedded, to edge, to cloud. It can be used to build streaming AI pipelines for a variety of domains, including medical devices, high-performance computing at the edge, industrial inspection and more.
NVIDIA Jetson is a platform for robotics and embedded edge AI applications. It offers compact, yet high-performance compute, supported by the NVIDIA JetPack software development kit for accelerated software development and NVIDIA application frameworks like NVIDIA Holoscan, NVIDIA Metropolis and NVIDIA Isaac for faster time to market.
“Humanoid robotic systems are complex and demand high processing capabilities for AI. OMNIVISION’s total camera solution supports four cameras simultaneously capturing high-resolution color images with an integrated ISP, which serves as a backend resource to support more AI functions. The solution has been verified and is ready with the NVIDIA Holoscan sensor bridge, which features an Ethernet output and supports long-distance transmission for flexible robotic camera placement,” said Kelly Yan, senior product marketing manager, OMNIVISION. “Our roadmap for robotics includes bringing an even more compact design to market in the future, using our recently announced 2.2-micron GS product family.”
OMNIVISION’s high-performance 2-megapixel (MP) OG02B10 GS sensor uses advanced 3 x 3-micron OmniPixel®3-GS pixel technology. This technology eliminates motion artifacts and blurring and dramatically improves low-light sensitivity. OMNIVISION’s OAX4000 companion ISP is capable of processing up to four camera modules with 140 dB high dynamic range (HDR), reducing design complexity and space and improving overall reliability.
