On May 22, 2024, Shanghai Jinmai Electronic Technology Co., Ltd. (“Jinmai”), with the in-depth exploration of the intelligence of the automotive industry, relying on the full-stack technical capability and forward-looking accumulation of the combination of software and hardware, and based on ToF (time Flying method) 3D imaging technology, launched a spacious control full-stack solution to transform advanced science and technology and lead the new experience of intelligent interaction in the automobile cockpit.
Golden vein 3D ToF solution (for illustration only), ToF (Time of Flight) is 3D imaging technology, which continuously sends active light pulses to targets in a specific range and collects optical information such as phase offset produced by the reflected light signal of the object to judge the distance space and other three-dimensional information of the object.
Thanks to the maturity of sensors and AI, ToF, with its accurate distance measurement ability, has brought revolutionary changes to intelligent interaction and ushered in great application potential in the field of intelligent vehicles.
Jinmai 3D ToF solution has international advanced visual perception technology, higher dimensional interactive perception ability, fast response speed and high recognition accuracy.
Without physical contact, drivers can control the intelligent cockpit and third-party application function scenes by recognizing and analyzing facial expressions, lines of sight and gestures, such as 3D facial recognition B-pillar unlocking, face payment, phone calls, as well as intelligent operation of entertainment multimedia systems such as audio and video switching, volume adjustment, gesture games and so on.
Not only improve the convenience and safety of the driving process, but also inject more convenience and fun into the driving experience.
The conceptual solution attracted wide attention since its debut on 2023 Smart China Expo.
As a complete platform system-level solution, Jinmai is driven by the full stack of software and hardware technology, while deeply considering the compatibility and ecology of the solution, and is developed based on a mass-produced and scalable product platform.
The scheme is composed of depth camera module, solving algorithm, gesture 3D reconstruction algorithm and gesture control middleware, which not only simplifies the complexity of the system, but also realizes elastic networking, efficient development and rapid and flexible deployment.
it has the characteristics of seamless integration, intergenerational compatibility and flexible configuration.
The international advanced hardware performance, the best combination of engineering design and calibration, the depth camera standard module adopts Infineon’s ultra-inductive 3D REAL3 ™image sensor based on the fifth generation pixel technology, which can provide real-time depth data at millisecond capture speed, with the advantages of high integration, high speed and unique performance.
it is the world’s first gauge-level ToF sensor that meets the functional safety requirements of ISO 26262 ASIL-B.
The sensor’s patented background lighting suppression (SBI) technology is not affected by ambient lighting conditions and can work efficiently even at night or in low light, strong light, rain and snow, or dusty weather.
At the same time, the monocular system architecture, which does not require mechanical alignment and angle correction even in the event of vibration, drop or thermal bending, is the best choice for accurate and reliable depth data and the smallest camera module design.
The standard module of depth camera has relatively simple structure, good heat dissipation performance and higher reliability, and supports VGA high resolution 640 × 480, wide angle FOV110 °× 80 °, higher than 30fps high frame rate standard operation.
The 940nm standard laser module which meets the laser safety IEC60825-1 Class 1 level is adopted, and the RGB sensor supports 200w rgb images.
The master-slave mode can be compatible with the existing IR modules in the cabin, and can flexibly support the technical evaluation and development verification of various scenarios.
The “body” of the whole module is small and can be designed according to the installation environment and flexibly match the installation position in the cabin, such as vehicle rearview mirror, AB column, middle roof of the vehicle, under-screen structure and so on.
Based on Infineon’s leading image sensor characteristics and deep understanding of ToF technology, combined with its many years of mass production experience and industry-leading production platform development, Jinmai explores and obtains the optimal adaptation balance point between modulation pulse frequency, depth accuracy, imaging field of view and resolution under multiple scenes in the cabin, and optimizes, calibrates and compensates the hardware module through software and algorithm.
Give full play to the best 3D imaging and processing capability of the hardware module.
Jinmai 3D ToF standard module, “Development first, Ecology first”, creates a new benchmark for seamless integration across modes.
Jinmai is developed and designed with a unified architecture, unified software stack and unified standard interface to create a complete “full ecological, system-level” solution, which can seamlessly integrate Android, Linux, HarmonyOS, complete vehicle engine systems and any built-in applications, third-party applications and specific scenarios across platforms.
It takes as little as a week to achieve a standard Demo integrated deployment with OTA capabilities.
The software scheme is combined with mapping, filtering, gesture strategy, attitude strategy and other algorithms and deeply tuned.
The performance of the overall scheme is much better than that of similar solutions in the market, and high-precision and robust identification, manipulation and processing are realized.
Jinmai independently developed platform middleware to achieve maximum compatibility in algorithm adaptation, standard drive and application ecology.
The innovatively designed algorithm adaptation framework can quickly adapt to the algorithm scheme without manufacturers and shorten the development cycle.
there is an exclusive standardized driver framework for the bottom layer, which can be efficiently transplanted and reused on different hardware platforms.
for the upper application, the rapid deployment of logic to meet the needs of customers and the whole vehicle factory, as well as the flexible scheduling and development of various upper applications, and support more ecological application function expansion.
A deeply optimized and efficient middleware system platform can provide upper-level applications with recognition based on different parts of the human body and more functions, such as hand gestures, head posture, body sitting posture, as well as active security and intelligent sound effects.
Scan the key points of the hand to control a variety of gestures, such as moving, clicking, dragging, customizing gestures, etc.
support the combination of intelligent entertainment system and APP application to recommend appropriate audio and video programs / games for people of different ages or children.
support 3D facial recognition, and recommend different types of music or entertainment information under different emotions according to different passengers’ expressions or emotions.
According to the sitting posture of different passengers, the body posture and head posture strategy and response can be carried out, the seat angle can be adjusted, and even in case of emergency, the ejection angle and strength of the airbag can be adjusted in real time according to the sitting posture and position.
maximize the safety of passengers.
Qin Chen, deputy general manager of Jinmai 3D ToF system architecture, said: “Vehicular 3D ToF is a complex and highly integrated solution involving exposure.
Multiple software and hardware technical links such as units, optical imaging, various algorithms, system middleware, and module design severely test suppliers ‘full-stack development capabilities.
Based on its mature platform development concept for many years, Jinmai controls the reliability and security of solutions and products with advanced technology and mass production requirements throughout the entire process.
It delivers not only complete full-stack product solutions, but also end-to-end.
Full link technical services.
In the future, Jinmai will join hands with more ecological partners to deeply empower the implementation of multi-scene intelligent solutions in the cabin to jointly define and promote the new form of the next generation of smart cockpit.
” Return to First Electric Network Home>.
