Monocular camera distance estimation technology accelerates the future of ADAS and autonomous driving
Morpho, Inc. (Headquarters: Chiyoda-ku, Tokyo; President and CEO: Masaki Hiraga; TSE Growth Market: 3653; hereinafter “Morpho”) has begun offering “Morpho Distance Scanner™,” a new technology that enables distance estimation using just one RGB camera.
This technology contributes to the cost reduction and advanced functionality of automotive camera systems as a new distance estimation solution in the fields of Advanced Driver Assistance Systems (ADAS) and autonomous driving, serving as an alternative to LiDAR, stereo cameras, sonar, and millimeter-wave radar.
By leveraging cameras already integrated into vehicles, it enables the deployment of advanced distance estimation capabilities without the need for additional hardware. For OEMs and Tier 1 suppliers seeking to enhance ADAS functionality, this represents an easily deployable and immediately impactful solution.
In addition, this technology offers high scalability, with potential applications extending beyond the automotive domain, including smartphones and robotics.
Top left: Original video
Bottom: Distance estimation results (distance to the target displayed as a heat map).
Top right: Bird’s-eye view (a 15m x 20m bird’s-eye view created from the estimated distance for each pixel. The fact that the bird’s-eye view is not distorted indicates high accuracy of distance estimation. The blue areas are not visible in the original video.)
With the advancement of autonomous driving and ADAS, the number of sensors installed in vehicles has been steadily increasing year by year. Among them, accurately measuring the distance between the vehicle and its surroundings is a critical factor in ensuring safety. Traditionally, distance estimation has relied on LiDAR and stereo cameras. However, due to their high cost, equipping mid-range and lower-priced vehicles with such sensors has been challenging. In addition, these sensors often present installation space constraints and reduce flexibility in vehicle design.
To address these issues, Morpho has developed a distance estimation technology that operates solely with RGB camera through software. This innovation enhances safety functions while achieving both design flexibility and cost performance in vehicle development.
Morpho Distance Scanner™ is software that estimates the distance to objects from a single image captured by an RGB camera. It requires no additional hardware, making it possible to leverage existing automotive cameras as they are.
The evaluation was conducted using datasets corresponding to both short-range (1–4 m) and mid-to-long-range (5–30 m) distances. The graph below shows the mean relative error of distance estimation with respect to ground truth, based on the distance between the camera’s imaging plane and the target object.
For example, when the target distance is 10 m, the high-precision model yields a relative error of 3.8%, which corresponds to an average deviation of approximately 38 cm. Across the entire mid-to-long range (5–30 m), the mean relative error is 5.5%.
Moving forward, we aim to further enhance the datasets and refine model tuning, with the goal of maintaining consistently high accuracy across a wider variety of scenarios.
Morpho Distance Scanner™ is expected to be applied not only as a real-time distance estimation system for ADAS and autonomous driving, but also in a wide range of use cases, such as automatic summarization of in-vehicle video, annotation support, and accuracy enhancement through integration with other recognition models including detection and segmentation.
Looking ahead, the technology will also be expanded into more accessible domains such as smartphone applications, robotics, and surveillance cameras, thereby broadening the scope of distance estimation technology. Morpho will continue to evolve this technology beyond automotive applications, positioning it as part of the social infrastructure that enhances both safety and convenience.
Morpho, Inc. is a research and development company specializing in “image processing/AI (artificial intelligence).” The company globally provides advanced image processing technology as embedded software, primarily to smartphone and semiconductor manufacturers in Japan and overseas. Additionally, Morpho offers AI-powered image recognition technology that analyzes image information captured by cameras on edge devices or in the cloud for the automotive and industrial IoT sectors, realizing various innovations through cutting-edge imaging technology.
Morpho, Inc. – Public Relations: Sato
TEL: +81-90-1051-0138
Contact Form: https://www.morphoinc.com/contact
*Morpho and the Morpho logo are registered trademarks of Morpho, Inc.
Copyright© Morpho, Inc. All rights reserved.
