Cookie settings

Our website uses cookies to understand how people use our website in order for us to improve our online experience.

Neonode Depth Estimation

Automotive Innovation Using Monocular Depth Estimation

Automotive
MultiSensing
Depth Estimation
In-cabin Sensing

Monocular depth estimation brings advanced safety and intelligence to every vehicle. Neonode’s software-only solution delivers accurate 3D sensing with just one camera—enabling adaptive airbags, digital mirrors and intuitive gesture control. Innovation, made accessible for all.

In the rapidly evolving automotive industry, depth estimation technology is unlocking a new frontier of safety and convenience features, transforming how vehicles interact with their environments and occupants. From advanced driver assistance systems (ADAS) to innovative applications like Digital Mirror Augmentation, depth estimation enables vehicles to perceive spatial relationships within the cabin. This is made possible when a system is able to analyze image data from internal cameras to calculate the distance between objects and the camera itself, in turn creating a three-dimensional understanding of the environment.

For instance, Digital Mirror Augmentation tracks the drivers head position in order to show an image perspective that follows the drivers exact line of sight. Depth estimation also supports in-cabin monitoring, such as detecting occupant positioning and optimizing airbag deployment based on passenger proximity. And in future, it can also be adapted to new in-cabin sensing innovations – imagine immersive entertainment where the surround sound that knows exactly where the listeners ears are positioned, or dynamic seating that automatically adjusts the seat height based on the passenger’s length. With some imagination, the possibilities are endless.

Implementation of Depth Estimation in Modern Vehicles

Traditional depth estimation methods have often relied on specialized hardware, such as time-of-flight or stereo cameras, which introduce significant cost and complexity. This has slowed the adoption of this technology in mainstream vehicles. Neonode’s MultiSensing technology overcomes these barriers, offering a cost-efficient, software-driven solution that leverages a single in-cabin camera to deliver high-accuracy depth sensing, making advanced features accessible to a broader range of vehicles.

Time-of-flight
Cameras

Structured
Light

UWB
Radar

Stereo
Vision

Neonodes Monocular
Depth Estimation

Measures distance by timing light’s travel to an object and back.

Capture 3D shape by projecting light and analyzing pattern deformation.

Detects objects and distances using short, low-energy pulses across wide frequencies.

Estimates 3D environment by combining streams from two fixed cameras.

Neonode’s advanced machine learning algorithm extracts 3D position from a single 2D image.

Hardware
Cost

Low

High

Low

High

Low

Typical
Range

5 m

2 m

1-2 m

5 m

5 m

Computational
Need

Low

High

Low

High

Low

Robust to
Sunlight

x

x

Multi-purpose
Sensor

x

x

x

Low Power
Consumption

x

x

Compact
Sensor

x

x

Robust to
Reflective
Surfaces

x

x

x

Neonode’s Monocular Depth Estimation: A Software-Driven Revolution

Neonode’s monocular depth estimation technology stands out by utilizing a single, standard in-cabin camera, making it both cost-efficient and highly adaptable. Unlike traditional approaches that depend on multiple cameras or specialized sensors, Neonode achieves depth sensing entirely through advanced software algorithms. This breakthrough reduces hardware costs, simplifies integration and enables automakers to deploy advanced features across a wider range of vehicle models, from luxury to economy.

By leveraging a single camera, Neonode’s solution minimizes the computational and financial overhead, allowing automakers to implement cutting-edge features without significantly increasing production costs. This approach aligns with the industry’s push toward scalable, efficient technologies that meet stringent safety regulations and consumer expectations.

The Power of Synthetic Data and Optical Expertise

At the heart of Neonode’s monocular depth estimation is its innovative use of synthetic data, combined with decades of expertise in optical engineering. Synthetic computer-generated datasets mimic real-world scenarios and enables rapid and efficient training of neural networks. Unlike real-world data, which can be time-consuming and costly to collect and annotate, synthetic data allows developers to simulate diverse scenarios, such as varying lighting conditions, occupant behaviors, or object distances. This is particularly valuable for new innovations, where accurately annotating distances or angles in real-world images would be impossible.

Since 2017, Neonode has refined its synthetic data-driven approach, building neural networks capable of recognizing complex spatial relationships with high precision. For example, the technology can determine the exact distance of an occupant from the camera or the angle of their posture. By eliminating the need for extensive real-world data collection, Neonode accelerates development timelines, allowing automakers to bring new features to market faster.

Neonode’s expertise in optical engineering further enhances the technology’s performance. With a deep understanding of camera systems and light behavior, Neonode optimizes its algorithms to work seamlessly with standard in-cabin cameras, ensuring reliability across diverse conditions, such as low light or high glare. This combination of synthetic data and optical expertise positions Neonode as a leader in software-driven depth estimation.

Applications in Driver and In-Cabin Monitoring

Neonode’s monocular depth estimation serves as the foundation for a wide range of driver and in-cabin monitoring functionalities, addressing both safety and convenience. Key applications include:

Occupant Detection and Positioning
The technology accurately detects where and how occupants are seated, enabling features like adaptive airbag deployment or alerts for improper seatbelt routing.

Digital Mirror Augmentation
By providing precise depth information, Neonode’s solution enhances camera-based mirror systems, by moving the image in the digital screen to match the drivers line of sight.

Gesture and Behavior Recognition
The system can interpret occupant gestures or movements, enabling intuitive human-machine interfaces, such as gesture-controlled infotainment systems or driver attention monitoring.

Driver Monitoring
Utilizing depth sensing, the technology is the founding for Noenode’s GSR-compliant driver distraction and fatigue detection.

Neonode’s technology supports compliance with evolving safety standards, such as Euro NCAP protocols for occupant monitoring or UNECE regulations for cybersecurity and over-the-air updates, ensuring vehicles remain compliant in dynamic regulatory landscapes.

The Future of Depth Estimation in Automotive Innovation

As the automotive industry moves toward greater automation and connectivity, depth estimation will play an increasingly critical role in enabling safer and more adaptive vehicles. MultiSensing’s monocular depth estimation is at the forefront of this transformation, offering a scalable, cost-effective solution that meets the demands of modern vehicle design. By combining synthetic data, advanced algorithms and optical expertise, Neonode empowers automakers to innovate rapidly, delivering features that enhance safety, comply with regulations and elevate the driving experience.

With its proven track record and forward-thinking approach, Neonode is paving the way for a new era of automotive technology, where advanced features are no longer limited by hardware constraints. As the automotive industry continues to evolve, Neonode’s monocular depth estimation will remain a cornerstone of innovation, driving safer and smarter vehicles for the future.