As the automotive industry continues to add more imaging systems like surround view systems (SVS), camera monitoring systems, and e-mirrors to luxury and mainstream vehicles, not only is the demand for image sensors growing, it is also boosting the requirements for these devices. This includes the need for higher resolution, increased high dynamic range (HDR), LED flicker mitigation (LFM), and in some cases, better low light performance.
Automotive OEMs continue to drive sensor manufacturers to improve performance – pushing HDR up to 140 dB, LFM that operates over the full temperature range, and better low-light performance. But there are always trade-offs among requirements such as resolution and low-light performance as well as cost and pixel size.
OmniVision has been able to meet some of these challenges by developing new technologies such as Split-pixel LFM, Deep Well dual conversion gain, and Nyxel near infrared (NIR), positioning itself to meet the growing demands of automotive OEMs.
The company has been supplying automotive image sensors since about 2005. It targets six market segments of which mobile, security, and automotive are the largest. These market segments call for a range of products, from the smallest packages for mobile devices to the lowest power imagers for security products.
One of the advantages that OmniVision offers is that after it develops a technology for a specific market, it can then apply it to other segments. One example is the introduction of its Nyxel technology, originally developed for the security market. Now it’s being applied to automotive applications.
What does it take to make cars see better? It’s complicated. Multiple sensing modalities, faster processing power in the vehicle brain, new car architecture with updated in-vehicle network, self-cleaning sensors, common sense layer and a lot more.
HDR and LFM pain points
HDR and LFM are critical key performance indicators (KPIs) for image sensors, especially for the OX08B40 and OX03C10 devices, said Mathew Arcoleo, product marketing manager, at OmniVision, during his presentation at AutoSens 2020. Both devices use OmniVision’s HALE (HDR and LFM engine) combination algorithm to provide HDR and LFM performance simultaneously.
LED flicker mitigation exists in a wide variety of situations, including headlights, billboards, road signs, just to mention few, said Arcoleo.
So, why are LEDs an issue? “LEDs are not always on, but are pulsed to control the power as well as the brightness,” said Arcoleo. “If the image sensor captures an image of an LED it could be on or off and this is what causes flicker. The high dynamic range is important with scenes with high contrast, like exiting a tunnel or looking directly into sunlight.”
“HDR is required when there is both bright and dim areas within the single image,” he said. “An example is driving in a tunnel where it’s relatively dark within the tunnel, but as you exit, it’s quite bright as you get into the sunshine.”
Although OmniVision’s OX08B40 (for high resolution front-view) and OX03C10 (surround-view as well as e-mirrors) parts target different applications, both require high HDR and LFM simultaneously. Both imagers utilize the PureCel stacked die technology that delivers a number of benefits, including higher dynamic range and low power, which is said to be about 25% lower than the competitions.
The capture schemes used for the LFM and HDR show that the devices will achieve 140 dB HDR and LFM simultaneously, which aligns with the customer’s KPIs, said Arcoleo. In addition, both offer ultra-low-power consumption and meet the ASIL-C automotive safety standard.
The OX08B40 is the industry’s first 8.3-megapixel (MP) sensor platform with optimized LFM and HDR for autonomous driving. In order to achieve the required resolution for the front-view application a small 2.1-micron pixel was developed for the OX08B40.
“It is the industry’s first 8.3-MP sensor with LFM and it’s also the first 8.3-MP sensor using the optimal HDR scheme. It is low power and pin compatible with other 8 megapixel variants.” The OX08B40 also has been approved by Nvidia for its Nvidia DRIVE AGX developer kit.
The OX03C10 is the first ASIL-C image sensor with a large 3.0-micron pixel size, 140 dB HDR, and LFM for viewing applications, delivering minimized motion artifacts. It’s also the first viewing image sensor with HDR and LFM that can deliver 1920 x 1280p resolution at the highest rate of 60 frames per second (fps) for faster camera-view switching for drivers. Automotive viewing applications include rearview cameras (RVC), surround view systems (SVS), camera monitoring systems (CMS), and e-mirrors.
The OX03C10 imager has the same HDR and LFM capture scheme as the OX08B40. “They differ in that the OX03C10 has a larger 3-micron pixel versus 2.1-micron pixel for the OX08B40. Both utilize the PureCel stacked die technology and have ultra-low-power consumption,” said Arcoleo.
The OX03C10 also leverages OmniVision’s Deep Well dual conversion gain technology to provide much lower motion artifacts and Split-pixel LFM technology with four captures for the best performance over the entire automotive temperature range. In addition, the PureCel Plus-S stacked architecture delivers greater pixel performance over non-stacked technology.
The combination of HDR, LFM, and large 3.0-micron pixel delivers the best image quality across all lighting conditions and in the presence of flickering LEDs from headlights, road signs, and traffic signals, according to OmniVision.
OmniVision and Geo Semiconductor Inc. recently announced a joint camera solution that delivers the capability to process images captured by two OX03C10 sensors with full HDR and LFM using a single Geo camera video processor. The image sensor takes advantage of the GEO GW5 CVP family’s ability to process 140 dB HDR images with full LFM at 60 fps, while the GW5 family’s advanced local tone mapping enables it to make optimal use of the OX03C10’s HDR and LFM image captures, translating into a more pleasing image for drivers.
Both companies’ devices provide advanced ASIL functional safety and AEC-Q100 Grade 2 certification for automotive applications.
High NIR QE
To meet growing demand for global shutter devices with high NIR quantum efficiency (QE), OmniVision offers the OV2312, claimed as the industry’s first RGB-IR sensor with global shutter for driver monitoring market (DMS)/in-cabin monitoring system (IMS) applications. The 2-MP device features low power and high NIR QE, and is compatible with earlier devices, namely the OV2310 and OV2311, said Arcoleo.
This automotive image sensor enables multiple functions, capturing both RGB and IR images, in one camera, which does a few things – reduces cost, space, and power budget for driver state monitoring (DSM) and viewing applications such as video conferencing for mainstream vehicles. It’s also small, housed in a 7.2 x 6.1 mm automotive chip-scale package, so cameras can be placed out of sight.
For operation without visible light, the OV2312 features the 3.0-micron OmniPixel3-GS architecture, which provides a NIR QE of 14% at the 940-nm wavelength, along with excellent modulation transfer function (MTF). This means the sensor captures images with the high quality required for driver eye and gaze tracking when running in single mode at 60 fps, while reducing system power consumption and cost by using fewer IR LEDs. The sensor only consumes 190 mW in typical conditions, reducing the heat generated for interior cameras used in tight spaces.
Earlier this year, OmniVision introduced its first automotive image sensor with Nyxel technology. OmniVision calls the Nyxel NIR technology a game changer for the DMS market, which requires low power, small size, and high NIR QE. The 2.5-MP OX03A2S rolling shutter RGB-IR sensor targets “short-range, light-starved automotive applications both inside and outside the vehicle,” said Arcoleo.
Other features include a 3.2-micron pixel size, ultra-small package (a-CSP package), low power consumption, best-in-class NIR sensitivity (40% at the 940-nm wavelength), and 120-dB HDR thanks to the Deep Well technology.
The NIR QE enables the sensor to detect and recognize objects that other image sensors miss under extremely low lighting conditions. The ASIL-B sensor uses RGB image captures during the daytime and monochrome at night and targets exterior imaging applications that operate in low to no ambient light conditions within two meters of the vehicle.
The Nyxel technology increases the QE of the device, and in this case [for the OX03A2S], the QE at 940 nm has increased from 8% to 40%, said Arcoleo. “This makes the OX03A2S ideal for light-starved applications both inside and outside the cabin. It also allows the OX03A2S to identify targets that other sensors simply can’t see.”
For a virtual demo on the four imagers, see OmniVision’s presentation at AutoSens 2020.
A new book, AspenCore Guide to Sensors in Automotive: Making Cars See and Think Ahead, with contributions from leading thinkers in the safety and automotive industries, heralds the industry’s progress and identifies the engineering community’s remaining challenges.
It’s available now at the EE Times bookstore.
Learn more about OmniVision Technologies