System and Method for Annotating Car Radar Data

Simple SummaryContent extracted from patent full text and abstract with AI.

This invention discloses a system and method for annotating car radar data by integrating data from at least one radar mounted on a vehicle and at least one optical detection system, such as a camera, positioned outside the vehicle. The camera produces images that are semantically segmented to identify different object classes (e.g., person, vehicle, bicycle) on a pixel level. Both the radar and camera data are aligned into a common coordinate system, allowing radar targets to be automatically labeled with the corresponding object class identified in the camera's segmented image. This approach enables efficient and accurate generation of annotated radar datasets, which are essential for training AI algorithms in autonomous vehicles.

Use CasesContent extracted from patent full text and abstract with AI.

• Automated labeling of radar data for use in machine learning and artificial intelligence systems for autonomous vehicles.
• Improving the quality of training datasets used to develop self-driving car perception algorithms.
• Enhanced annotation and classification of objects detected by automobile radar, even in complex environments or where objects overlap visually.
• Creation of public or proprietary high-quality radar datasets for research or industrial automotive purposes.
• Augmenting safety validation and testing processes for advanced driver-assistance systems (ADAS).
• Support for real-time object detection and classification in vehicles through improved radar data annotation.

BenefitsContent extracted from patent full text and abstract with AI.

• Significantly reduces the time and labor required for manual labeling of large radar datasets.
• Enables highly accurate and robust annotations by correlating radar targets with semantically segmented camera images.
• Allows detection and labeling of objects that are occluded or not directly visible from the vehicle's perspective, due to the external camera's vantage point.
• Facilitates the generation of superior training datasets, leading to better AI performance in object detection and classification for autonomous vehicles.
• Supports greater automation in the data annotation process, decreasing costs and annotation errors.
• Improves the exploitation of radar's full data potential, contributing to safer and more reliable autonomous driving systems.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

The invention relates to a system for annotating car radar data, comprising: at least one radar arranged on a car for producing a radar image by means of radar measurement; at least one optical detection system arranged outside the car for producing a camera image; a segmentation unit, which is designed to subject a camera image produced by the optical detection system to semantic segmentation for forming a semantic grid in order to assign one of a plurality of object classes to the camera image pixel by pixel; a computing unit, which is designed to transfer the camera image and/or the radar image into a common coordinate system for co-registration; and an annotation unit, which is designed to carry out annotation of the radar image, in other words to allocate an object class to a radar target of the radar image, in such a way that the object class of the semantic grid of the co-registered camera image in which the radar target of the co-registered radar image is located is allocated to a particular radar target.