NVIDIA NeuroEye

Intelligent Camera

NVIDIA Eye

NeuroEye is a high-definition AI Smart Camera designed based on the NVIDIA® Orin NX platform, featuring interchangeable lenses. It is equipped with Sony's high-definition and high-sensitivity sensor technology, supporting both rolling and global shutter sensors, facilitating easy capture across various application scenarios. With its independent hardware ISP design, video processing and optimization are conducted without consuming CPU/GPU resources. Additionally, with HDR (High Dynamic Range) support, it captures clearer and more detailed video in high-contrast scenes. The exclusive DMA PCIe Design High-Bandwidth Capture enables 4K60 RGB444 full-color video output. When coupled with YUAN's efficient AI QDEEP library, it showcases multiple AI applications simultaneously on the powerful NVIDIA platform. This high-spec Smart AI Camera is the optimal choice for commercial security, traffic monitoring, smart cities, and industrial applications.

 

— Broad Product Line —

  •  
 

 

— Feature —

Adopting Sony STARVIS 2 & Pregius S Sensor

NeuroEye uses Sony's latest STARVIS 2 technology image sensor series, enabling single-shot image capture with higher sensitivity and wider dynamic range. High quality images with minimal noise are captured even in low light conditions. In scenes with high-speed signals and high resolution, the Pregius S technology series supports image sensors with global shutters. This ensures clear images of fast-moving objects without motion blur, enabling more precise and accurate image data capture.

Hardware ISP Offers Outdoor HDR without CPU/GPU Load

The camera design with an external ISP allows separate processing of parameters like noise reduction, image sharpening, white balance, exposure, and dynamic range enhancement. This preserves processing power for more demanding tasks like imaging algorithms or parallel computation of AI models. This enhances overall system efficiency by enabling smoother handling of multiple functions.

DMA PCIe Design High-Bandwidth Capture Providing RGB444 Image Quality

The camera's integrated DMA PCIe design enables complete capture of high-quality, full-colour images. This unique design helps to reduce system latency, CPU utilisation, system power consumption and also improves system stability. Combined with the PCIe interface, it can handle large amounts of data efficiently. As a result, it reliably and efficiently captures 4K60 RGB444 full colour images with excellent stability and efficiency.

C-Mount Design for Versatile Lens Usage by Professionals

Designed with a C-Mount interface, this configuration allows for flexible interchangeability of lenses to accommodate various focal lengths and aperture requirements according to specific application scenarios. Additionally, it is more suitable for larger-sized image sensors, enabling the capture of more light sources and thus enhancing image quality. This setup is particularly advantageous for applications in industrial vision and machine vision, offering a flexible and cost-effective solution.

ISP Tuning Tool

NexCam Studio is a dedicated ISP adjustment tool developed specifically for cameras. It encompasses both basic and advanced image parameter settings, including white balance, shutter speed, brightness, contrast, etc. Additionally, it incorporates image processing functionalities such as cropping and flipping. With backend image snapshot saving, it offers flexible configurations for real-world applications, enabling quick and easy handling of various usage scenarios.

All in One Design

NexCam Studio is an ISP adjustment tool designed specifically for cameras. It includes basic and advanced image parameter settings, including white balance, shutter speed, brightness, contrast and more. It also includes image editing functions such as cropping and mirroring. With its back-end image capture function, it provides flexible configurations for real-world applications, enabling real-time and powerful support for different application scenarios.

— Technology —

 

Generative AI

Equipped with exceptional computing power and ample memory, Jetson Orin is particularly well-suited for Generative AI applications. Generative AI is better suited to different scenarios than discriminative AI in the past. Unsupervised or semi-supervised learning allows Generative AI to better understand hidden information in videos, making it easier to recognize and locate objects in different scenes without the need for repeated labeling and training.

Advanced AV1 4:4:4 Codec Core

The VPP6N0 adopts the industry-leading AV1 4:4:4 encoding and decoding technology, ensuring minimal compression loss while preserving more original video details. It also offers higher transmission speed and clarity.

NVIDIA® GPUDirect® and GPUDirect RDMA Technology

YUAN's VPP6N0 series fully supports NVIDIA® GPUDirect® and GPUDirect® Remote Direct Memory Access ( RDMA ) technologies. This integration significantly improves image access performance and reduces CPU loading, meeting the demands of ultra-low latency video applications.

 

Exclusive 5G Low-Latency Streaming Technology

Built on the Orin NX platform, VPP6N0 integrates a multi-channel 5G module design for long-range video transmission. With 4K60 4:4:4 video capture capability, advanced H26X and AV1 encoding, and YUAN's SkyLink X streaming, it achieves ultra-low latency ( ≦ 50ms ) from capture to display. VPP6N0 is ideal for remote control applications like drones and autonomous vehicles.

— Application —