A vehicle four-way fisheye camera panoramic splicing method

The panoramic stitching method using four fisheye cameras for vehicles, based on the MSER algorithm and state switching model, solves the problems of insufficient field of view and poor stitching effect in existing technologies, achieving a wide field of view and efficient stitching, and is suitable for the field of autonomous driving.

CN115936995BActive Publication Date: 2026-06-19SOUTH CHINA UNIV OF TECH
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SOUTH CHINA UNIV OF TECH
Filing Date
2023-01-03
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

Existing in-vehicle 360-degree panoramic surround view systems cannot provide a wide field of vision when the car is moving forward, especially the blind spot information of the A-pillar. In addition, the existing fisheye camera calibration and correction methods have decoupling problems, resulting in poor stitching effect and slow stitching speed.

Method used

The MSER algorithm is used for fisheye image calibration and distortion correction. The cylindrical projection model or perspective transformation model is switched according to the vehicle driving state for image stitching. The stitching seam is optimized using the mask cropping method and the distance trigonometric function weighted average fusion algorithm to generate a panoramic stitched image.

Benefits of technology

It provides panoramic information that switches according to the vehicle's driving status, improving the field of view, reducing computational load, minimizing stitching seams, meeting the needs of autonomous driving, and improving the stitching effect.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115936995B_ABST
    Figure CN115936995B_ABST
Patent Text Reader

Abstract

This invention discloses a panoramic stitching method for a vehicle using four fisheye cameras, comprising the following steps: 1) acquiring images from fisheye cameras installed around the vehicle and the vehicle's gear position signal; 2) using the MSER (Most Stable Limit Region) algorithm for calibration and image distortion correction; 3) determining the vehicle's driving state based on the gear position; 4) if in forward gear, using a cylindrical projection model to obtain a wider field of view, and if in reverse gear, using perspective transformation to obtain a bird's-eye view of the vehicle; 5) stitching using a mask clipping method; 6) obtaining a panoramic stitched image of the vehicle's surroundings in forward gear and a bird's-eye view panoramic image of the vehicle's surroundings in reverse gear; 7) optimizing the image stitching seams using a fusion algorithm; and 8) generating a panoramic stitched image. This invention allows switching between different stitching modes based on the vehicle's driving state, providing the driver with a better field of view. It has the advantages of short processing time, high accuracy, good results, and ease of implementation.
Need to check novelty before this filing date? Find Prior Art