author: Eric Walz
StradVision is developing advanced perception systems for autonomous vehicles and ADAS.
Advances in computer graphics and AI-powered processing have led to the development of more robust perception systems for advanced driver assist systems (ADAS) and for autonomous driving. The world’s automakers are forming partnerships with companies that are working on these technologies for the next generation vehicles that are being equipped with advanced safety features.
One such company working on the technology is StradVision. The South Korea-based company is a pioneer in AI-based vision processing technology for vehicle ADAS and autonomous driving and has just announced two major partnerships with a leading German automotive OEM and Tier-1 supplier.
StradVision will supply its augmented reality technology for Navigation and Lane Keeping Assistance Systems (LKAS). The second partnership is with a leading global Tier-1 supplier to provide a Surround View Monitoring (SVM) algorithm that supports Park-Assist functions such as Automatic Parking Assistance (APA).
Surround View Monitoring and Auto-Valet features are being introduced on many new vehicle models today.
StradVision did not name the companies, but said its augmented reality technology will be featured on more than 40 vehicles from the OEM, the company said. The vehicles are scheduled for production globally in 2022.
StradVision’s SVNet Deep-Learning Perception Software
StradVision’s computer vision and AR technology is powered by its SVNet deep-learning-based perception software, which supports enhanced depth-mapping solutions, semantic segmentation, and lane detection for lane-keep assist and other ADAS safety features. SVNet enables vehicles to execute ADAS and self-driving functions, such as automatic emergency braking and adaptive cruise control by improving perception.
SVNet is one of the few novel networks in the industry that meets the accuracy and computational requirements for commercial use on automotive embedded hardware, the company says.
StradVision’s augmented reality vision solutions are able to solve issues related to perception, including occlusion, where the object being tracked is obscured by another object in the foreground. SVNet is also being used to improve the positioning and pitch compensation, which keeps AR elements fixed to their proper positions in a 3D simulated space. This technology can support navigation maps that appear in 3D.
“Drivers of high-end vehicles have grown to expect the best accuracy, performance and visual quality from their in-vehicle experiences, and StradVision’s technology can take their expectations to the next level,” said StradVision CEO Junhwan Kim. “Our advanced approach to providing augmented reality for these vehicles will be rightfully recognized as a major leap forward.”
Surround View Monitoring and Automatic Parking Assistance
StradVision’s Surround View Monitoring solution is one that the company believes will be in high-demand for many years. Surround view monitoring can mimic how a vehicle looks from above using images stitched together from a vehicle’s cameras. It’s typically used as a parking aid.
A similar feature is founded on the Volvo XC90 SUV. Volvo calls its “360° Surround View”, but its commonly referred to as a “bird’s-eye-view.”
Using computer vision techniques and fish eye cameras lenses, Volvo’s surround view system mimics how the vehicle would look from above to assist a driver when parking, backing up, or maneuvering out of a space. However, there is no actual camera above the vehicle, the scene entirely simulated using software, augmented reality and computer vision processing.
StradVision’s Surround Vision technology can also be used by vehicles that are capable of self-parking. The company has begun working with a global Tier 1 automotive supplier to provide a Surround View Monitoring (SVM) algorithm that supports automated park-assist functions.
“Imagine the peace of mind that will come with knowing you don’t have to park in a difficult spot for yourself, and can let the vehicle do the work for you,” Kim said. ” As the industry continues to trend toward self-parking vehicles and Auto-Valet features, StradVision’s vision processing expertise will be extremely helpful to Tier 1 suppliers and OEMs making this transition. We intend to be a leader in this field, and are excited to be ahead of the curve in making people’s lives easier with this technology.”
Surround view monitoring technology can be used as a parking aid, including for vehicles capable of self-parking.
Since launching in 2014, StradVision’s deep-learning, AI-based expertise has helped the company become a leading developer of software for object detection (OD) and free space detection (FSD) around a vehicle.
StradVisions’s SVNet software also enables vehicles to execute ADAS and self-driving functions for features such as Automatic Emergency Braking (AEB), Adaptive Cruise Control (ACC), Lane Departure Warning (LDW) and Blind Spot Detection (BSD).
To date, StradVision’s SVNet is deployed on 50+ vehicle models in partnership with 9 OEMs and powers 13 million ADAS & Autonomous Vehicles worldwide, the company said.
In 2020, StradVision received the Autonomous Vehicle Technology ACES Award in the autonomy software category. The award recognizes innovative technologies in the areas of vehicle autonomy, connectivity, electrification, and mobility services (ACES).
StradVision is currently engaged in multiple production projects in Europe and China, including work on a perception system for a Level 4 autonomous bus project in Europe.