Method for detecting an object and motor vehicle
The method employs a search pattern of curves from the vehicle to assign pixels across overlapping image regions, improving real-time object detection accuracy and reducing data volume, addressing the challenges of existing pixel assignment methods.
DE102024101247A1Pending Publication Date: 2025-07-17CARIAD SE +1
4 Cites 0 Cited by
Patent Information
- Application Number
- DE102024101247
- Authority / Receiving Office
- DE · DE
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-01-17
- Publication Date
- 2025-07-17
AI Technical Summary
Technical Problem
Existing methods for detecting objects using multiple image sensors on a vehicle struggle with efficiently assigning pixels from overlapping image regions in real-time, leading to difficulties in accurate and timely object recognition.
Method used
A method involving the use of a search pattern of curves originating from the vehicle, such as radial beams or branches, to assign pixels across overlapping image regions, combined with time stamps and vehicle motion data, to facilitate precise object detection.
Benefits of technology
Enables reliable and accurate real-time object detection by reducing data volume and enhancing pixel assignment accuracy, even with asynchronous image capture, using a spatially fixed grid pattern.
✦ Generated by Eureka AI based on patent content.
Abstract
To detect an object based on camera images from cameras (12, 14) with overlapping image areas, pixel groups in the respective images should be able to be assigned. For this purpose, a system that is world-fixed with respect to the cameras and has curves (L1, L2, L3, ..., L48) emanating from the camera carrier is used. This is particularly useful in the automated driving of a motor vehicle (1) as the camera carrier and also when assisting the driver.
Need to check novelty before this filing date? Find Prior Art