Ambarella, a leading developer of high-resolution video processing and computer vision SoCs, has partnered with Neusoft Reach Automotive Technology to produce advanced driver assistance system (ADAS) smart cameras for Chinese original equipment manufacturers (OEMs). The companies have already launched a third-generation X-Cube 3.0 front view smart camera incorporating Ambarella’s CV22 artificial intelligence (AI) vision system-on-chip (SoC).
Mass Production of Front-ADAS Smart Cameras
Neusoft Reach has started mass-producing the X-Cube 3.0 smart camera, which is being deployed in passenger cars to improve safety and efficiency. The camera relies on the CV22 SoC to run the L2 AI perception algorithm, which allows it to detect and respond to potential road hazards. The X-Cube 3.0 smart camera has also been integrated into various vehicle models that entered series production in 2022.
Neusoft Reach’s L2 Front-ADAS Smart Camera
The third-generation X-Cube 3.0 smart camera developed by Neusoft Reach features a compact design and can deliver high-quality video feeds. It comes with a host of features that allow it to analyze road conditions and detect obstacles in real time. The camera can also recognize traffic signs, traffic lights, and pedestrians, making it a valuable safety asset for passenger cars.
The camera’s L2 AI perception algorithm is optimized for automotive applications, making it ideal for use in front ADAS systems. The algorithm uses a neural network to identify objects and predict their movements, allowing the camera to alert the driver to potential hazards. The algorithm is also capable of distinguishing between stationary and moving objects, reducing the likelihood of false alarms.
Ambarella’s CV22 AI Vision SoC
Ambarella’s CV22 AI vision SoC is a powerful processor designed to handle high-resolution video and image processing tasks. The SoC features an advanced image signal processor (ISP) that can capture high-quality images in a range of lighting conditions. The ISP is also capable of reducing noise and improving color accuracy, resulting in clearer images.
The CV22 SoC is also equipped with Ambarella’s CVflow architecture, which is optimized for deep neural network processing. This allows the SoC to deliver high-performance AI processing while consuming minimal power. The CV22 SoC can also support multiple neural networks simultaneously, allowing for complex ADAS applications.
Conclusion
The partnership between Ambarella and Neusoft Reach has resulted in the development of advanced front ADAS smart camera products that are now in mass production for Chinese OEMs. The X-Cube 3.0 smart camera, based on Ambarella’s CV22 AI vision SoC, is capable of delivering high-quality video feeds and analyzing road conditions in real time. The camera’s L2 AI perception algorithm is optimized for automotive applications, making it a valuable safety asset for passenger cars. The CV22 SoC’s advanced image signal processor and efficient AI processing capabilities make it a powerful solution for a range of ADAS applications.
