Perception system for autonomous vehicles

By installing cameras on autonomous vehicles and using image processing technology to crop regions of interest, combined with machine learning and topographic mapping, the challenge of detecting objects at long distances in autonomous vehicles has been solved, enabling 3D localization of objects at long distances and safe driving.

CN113838060BActive Publication Date: 2026-06-05ORIGINAL GENERATION CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
ORIGINAL GENERATION CO LTD
Filing Date
2021-06-22
Publication Date
2026-06-05

AI Technical Summary

Technical Problem

Autonomous vehicles have limited perception range, especially semi-trailer trucks, making it difficult to safely detect and avoid objects on the road at long distances, leading to potential collision risks.

Method used

By installing cameras on autonomous vehicles, using image processing techniques to crop regions of interest, and combining machine learning and topographic mapping, the location of objects can be detected and determined, enabling 3D localization and motion model tracking of distant objects.

Benefits of technology

It improves the safe driving capabilities of autonomous vehicles, enabling them to accurately identify and avoid objects at long distances, thus enhancing the safety of autonomous driving on roads.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113838060B_ABST
    Figure CN113838060B_ABST
Patent Text Reader

Abstract

The present disclosure relates to a perception system for an autonomous vehicle. Image processing techniques are described to perform the following steps: obtaining an image from a camera located on the vehicle while the vehicle is being driven; cropping a portion of the obtained image corresponding to a region of interest; detecting an object in the cropped portion; adding a bounding box around the detected object; determining a location of a reference point on the bounding box; and determining a location of the detected object in a spatial region in which the vehicle is being driven based on the determined location of the reference point on the bounding box.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • System and method for image localization based on semantic segmentation

    US20180336421A1

  • System and method for actively selecting and labeling images for semantic segmentation

    US20180365835A1

  • Three-dimensional bounding box from two-dimensional image and point cloud data

    CN111108507A

  • Method for determining a current distance and / or a current speed of a target object based on a reference point in a camera image, camera system and motor vehicle

    US20150035973A1