Ruisi Zhixin completes nearly 100 million yuan of Pre-A round of financing, focusing on a new generation of machine vision technology

On July 7th, Beijing Ruisi Zhixin Technology Co., Ltd. (referred to as "Ruisi Zhixin") has completed nearly 100 million yuan of Pre-A round of financing, jointly led by Hikvision and Yaotu Capital, iFLYTEK Venture Capital, Sunny Central Research Institute, Allwinner Technology, Chasing Venture Capital, Tongchuangweiye followed suit, and old shareholders Lenovo Venture Capital and Zhongke Chuangxing continued to raise their stakes.

Founded in 2019, Ruisi Zhixin is a vision sensor technology company headquartered in Beijing, with branches in Shenzhen and Switzerland. It has complete core intellectual property rights and independent research and development capabilities, which can fill the technology and patents in the domestic machine vision field. blank.

With the core of Hybrid Vision technology, the company has groundbreakingly created fusion bionic vision chips and integrated machine vision solutions and is committed to continuing breaking through the development bottleneck of the industry and empowering the entire machine vision ecosystem. Hybrid Vision technology came into being in response to the bottleneck of the development of traditional vision sensors. It is a new generation of machine vision technology that combines the advantages of bionic vision and traditional image sensors.

Although the traditional image sensor has high-quality imaging, the design source is to adapt to the human eye rather than the machine. Its image acquisition method is based on a fixed frequency "frame". There is a lack of information between frames, which makes continuous recording impossible. The machine's large amount of data is repeated and redundant, and its exposure operation method also determines the short dynamic range and cannot adapt to the defects of extreme lighting environments. Therefore, it does not have the fast response, low system delay, and high environmental adaptability required by machine vision.

The bionic vision sensor is driven by events and imitates the working principle of the human eye's retinal neurons. It records and outputs event signals with a very fast response speed only when changes are sensed, so as to achieve the continuous perception of scene changes with low power consumption. Effect; At the same time, the parameters can be automatically adjusted to adapt to various lighting conditions.

Hybrid Vision-fusion bionic vision technology, which integrates bionic vision technology and traditional image sensors. It can imitate the working principle of human retinal neurons, drive based on events, and capture changes in the full-time domain with a response speed of microseconds. It also inherits the technical advantages of traditional image sensors, guarantees the full-frame image quality and image details in the field of image shooting, so as to achieve the effect of simultaneously outputting high-quality images and event stream data from one chip.

This technology eliminates redundant data from the source, greatly reduces the amount of system data processing, and reduces system costs; and automatically adjusts parameters to adapt to the current lighting conditions to achieve a high dynamic range, ensuring extreme and rapid changes in light and dark lighting conditions Maintain good operation; at the same time, it is the same as the traditional image sensor manufacturing process, compatible with existing image processing algorithms and architecture, and can be quickly adapted and applied to actual work scenarios.

Under the innovation derived from independent patented technology, the Hybrid Vision-a sensor empowered by fusion bionic vision technology enables the machine to have lower system delay, faster response-ability, and better dark light performance, A higher signal-to-noise ratio, and a series of advantages such as lower cost.

At present, Ruisi Zhixin has launched its core product-a new generation of bionic vision sensor chip ALPIX, which can be widely used in machine vision scenarios in the fields of smart home, security monitoring, 3C electronics, and autonomous driving.

For smart homes, ALPIX relies on the low data redundancy of its event stream data, the efficiency of capturing dynamic changes, and privacy protection, reducing system processing costs, avoiding omissions and misjudgments, and realizing low-power smart triggers and intelligence. Tracking and monitoring functions in a privacy environment. Based on the high dynamic range advantage of ALPIX event stream data, it can realize the normal operation of equipment in all-weather, including non-light conditions.

In terms of autonomous driving, the ALPIX fusion solution can perfectly cooperate with mainstream sensors such as TOF to build a complete autonomous driving vision system to achieve finer image judgment, more accurate distance measurement, and higher-speed real-time prediction functions. Based on the low data redundancy of the ALPIX event stream data, the system processing cost is reduced, and the automatic driving low-latency research and judgment function is realized, so as to quickly respond to road emergencies. Based on the high dynamic range advantage of the ALPIX event stream data, it realizes the high environmental adaptability of autonomous driving and avoids the visual blind spot phenomenon that occurs under extreme lighting conditions or rapid changes in light and darkness.

In terms of security monitoring, ALPIX relies on the high dynamic range advantage of its event stream data to achieve clear shooting in low light conditions. Based on the high efficiency of the ALPIX event stream data to capture dynamic changes, it realizes the intelligent judgment and tracking function of specific characteristics, in the end, test, as well as the rapid identification and retrieval of characteristic information.

Ruisi Zhixin has a team with both algorithm, software, hardware, chip research, and development experience and capabilities. The core experts are all from the world’s top scientific research companies and institutions, including CSEM, NXP, ARM, Freescale, ON Semiconductor, Magic Leap, Intel, etc., have vertical industry experience in the development of multiple and multiple generations of commercial chips.

In terms of intellectual property rights, Ruisi Zhixin has mastered core intellectual property patents in many aspects from principle, design; implementation, application, etc., including 8 EU patents and 4 countries in the fields of pixel design, chip, and system architecture, and algorithms. Invention patents and 1 PCT.

Different from other DVS technology paths, RISC uses an innovative pixel structure to design its own machine vision sensor chip, which can realize DVS and RGB image functions on the same sensor chip, which has more advantages in function and performance.

Artificial intelligence is profoundly changing all walks of life, and the digitalization and intelligence of the industry will be the subject of industrial transformation in the next few years. New sensors can provide richer data for artificial intelligence technology. In the field of vision, DVS relies on its significant differences and advantages to form a good complement to traditional image sensors and has a wealth of application imagination.

It can be seen that these are also the reasons why Hikvision and iFlytek are optimistic about and invest in Ruisi Zhixin. Xunfei Ventures stated that new sensors have always been the focus of the company focus, and Ruisi Zhixin has initially implemented DVS sensor products. In the future, the company will also use its own algorithmic advantages to explore various new applications with Ruisi Zhixin to better facilitate the digital and intelligent transformation of the industry.