As people's expectations for travel safety continue to rise, the deployment of ADAS (Advanced Driver Assistance Systems) is becoming increasingly widespread. From early, simple reversing radars to today's complex and sophisticated functions such as automatic emergency braking and adaptive cruise control, ADAS is providing comprehensive and multi-layered protection for driving safety. Among them, the core foundation of camera-based ADAS systems is the image sensor. Common rearview cameras allow drivers to clearly see the situation behind the vehicle, effectively reducing the risk of collisions when reversing; 360-degree surround view systems, by cleverly installing multiple cameras around the vehicle, present the driver with a panoramic view of the vehicle's surroundings, showing significant advantages in low-speed parking and maneuvering scenarios, greatly improving parking safety and convenience. Forward-facing camera systems not only show the driver the road conditions ahead in real time, but also provide core data support for key functions such as automatic emergency braking and adaptive cruise control by accurately sensing the dynamics in front of the vehicle, helping the vehicle make quick and accurate decisions in complex road conditions and effectively avoid collisions.
Different automotive camera applications place vastly different demands on image sensor performance. For example, rearview cameras, used to intuitively display images to the driver, focus on providing clear and easily understandable images, allowing the driver to quickly grasp the situation behind the vehicle. Frontview cameras, as key "sensory organs" of ADAS systems, require a higher level of image quality. They must not only clearly present objects ahead but also, combined with complex algorithms, accurately identify pedestrians, vehicles, and other obstacles, laying a solid foundation for safety functions such as automatic emergency braking. From a resolution perspective, the resolution requirements of image sensors used for specific display formats differ significantly from those required for computer vision applications. Computer vision applications rely on a precise minimum number of pixels for algorithms to correctly detect and identify objects, which undoubtedly places stringent demands on sensor resolution. Taking highway scenarios as an example, to accurately identify distant traffic signs and vehicles in advance, the image sensor of the frontview camera needs to have high resolution, providing sufficient detail information to ensure the ADAS system can react effectively in advance.
To meet the diverse and stringent requirements of automotive safety features for image sensors, major manufacturers have increased their R&D investment and actively launched advanced image sensor platforms. ON Semiconductor's Hayabusa™ automotive image sensor platform stands out as a leader in this field. This platform innovatively employs pixel technology that enables Super-Exposure, successfully overcoming two key challenges in automotive imaging: high dynamic range imaging and LED flicker suppression. Through unique design and manufacturing processes, each back-illuminated 3.0-micron pixel of the Hayabusa™ image sensor can store over 100,000 charged electrons, a significant leap in charge storage compared to traditional CMOS image sensors. This allows for an exposure time five times longer than traditional sensors before saturation. This superior characteristic enables the sensor to achieve high dynamic range imaging exceeding 120 dB, easily handling the complex and varied lighting conditions encountered during vehicle operation. Whether on a highway under direct sunlight or in a dimly lit underground parking lot, it clearly captures details in both bright and dark areas, effectively avoiding overexposure or underexposure, providing accurate and reliable visual information for ADAS systems. Meanwhile, when dealing with the flickering problem caused by pulsed LED light sources such as traffic lights and LED vehicle lights, the Hayabusa™ platform cleverly limits the super exposure time to ensure that the entire cycle of the lowest frequency pulsed LED in the scene can be captured. Combined with on-chip algorithms and secondary short exposures, it effectively suppresses LED flicker, ensures the stability and accuracy of imaging, and significantly reduces the risk of misjudgment caused by image flicker in ADAS systems.
OmniVision's OX08D10 image sensor has also shone brightly in the automotive vision field. As the first product to adopt the new 2.1-micron TheiaCel technology, it successfully eliminates the interference of LED flicker on imaging under various lighting conditions by utilizing next-generation lateral overflow integral capacitor (LOFIC) and DCG high dynamic range (HDR) technology. This is significant for accurately recognizing traffic signs and signal lights, greatly improving the reliability of autonomous driving and driver assistance systems in complex lighting environments. Moreover, thanks to this technology, the OX08D10 can achieve HDR image capture at distances up to 200 meters, achieving an excellent balance between signal-to-noise ratio and dynamic range. This provides a wide field of view and clear images for exterior camera applications, allowing drivers and autonomous driving systems to perceive potential hazards in advance, gaining more valuable reaction time for safe driving.
SmartSens' SC360AT, a 3MP high-performance automotive-grade image sensor, is meticulously crafted using CarSens®-XR Gen 2 technology and employs a Stacked BSI + Rolling Shutter architecture. This sensor incorporates advanced technologies such as SuperPixGain HDR™ 2.0 and LFS, boasting outstanding advantages including high sensitivity, high dynamic range (up to 140dB), low noise, and low power consumption, while also supporting LED flicker suppression. In extreme lighting conditions such as backlighting and strong light, including the intense contrast at tunnel exits and the illumination of oncoming headlights at night, it effectively addresses the challenges of overexposure in bright areas and loss of detail in dark areas, ensuring clear and accurate real-time images for automotive cameras. This provides robust support for various ADAS applications, including side-view, rear-view, and surround-view imaging. Furthermore, it complies with both ISO 26262 ASIL-B automotive functional safety and ISO 21434 automotive cybersecurity standards, leveraging high reliability and safety to accelerate the transformation of automotive intelligence and connectivity.
These advanced image sensor platforms not only excel in technical performance but also bring significant benefits to automakers in practical applications. They typically possess excellent scalability, encompassing a variety of resolutions from low to high pixel counts. Manufacturers can flexibly select appropriate image sensors based on different vehicle positioning and ADAS functional requirements, while reusing engineering development work and image training datasets, significantly reducing R&D costs and time. For example, high-resolution, high-performance image sensors can be used in high-end models to achieve higher levels of autonomous driving functions, while relatively low-pixel but cost-effective products can be used in mid-to-low-end models to meet basic safety assistance needs. This scalability enables automakers to quickly deploy diverse ADAS systems on different vehicle platforms, accelerating the popularization and upgrading of automotive safety features.
Image sensor platforms are accelerating the deployment of automotive safety features across the board thanks to their outstanding technological innovations. From improving image quality and mitigating complex lighting and LED flicker issues to providing manufacturers with scalable, low-cost solutions, image sensor platforms are injecting powerful momentum into the development of ADAS, bringing safer and more reliable travel protection for drivers and pedestrians. With continued technological advancements, image sensors are poised to play an even more crucial role in future automotive safety, propelling autonomous driving technology to new heights and leading a new trend in automotive safety development.
