Surround View Image Stabilization via Vanishing Point Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current driver assistance systems face challenges in accurately stabilizing images from vehicle cameras, particularly in surround view systems, due to lens distortions and camera motion, which can introduce errors and require computationally intensive methods.
Innovation Solution
The method involves detecting parallel lines, calculating vanishing points on a unit sphere, tracking their motion using a prediction filter like a Kalman filter, and applying lens distortion corrections to project images to a ground plane, thereby stabilizing the image sequence and correcting for camera orientation changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional image stabilization methods (block motion vectors, SIFT matching) are used, then motion estimation can be achieved, but errors are introduced (blocks appearing, computational intensity) and reliability decreases
Solution Approach 1:
The patent extracts and utilizes vanishing points from the image sequence as key features for motion estimation. By focusing on these specific geometric features rather than general feature matching or block-based methods, the system achieves more reliable stabilization with reduced computational overhead. The vanishing points are detected from straight lines in the corrected images and used to derive camera motion directly.
Solution Approach 2:
The patent projects image points onto a unit sphere to calculate vanishing points, transforming the 2D image coordinate problem into a 3D spherical coordinate problem. This dimensional transformation enables more accurate representation of camera motion and improves the reliability of motion estimation while maintaining computational efficiency through geometric relationships on the sphere.
2Manufacturing precision
If lens distortion corrections are applied to straighten lines, then manufacturing precision of image geometry is improved, but device complexity increases due to additional processing steps
Solution Approach 1:
The patent applies lens distortion correction as a preliminary step before vanishing point detection and motion estimation. By pre-correcting the images to remove barrel distortions and straighten lines, the subsequent geometric processing becomes more accurate and reliable. This preliminary action ensures that the vanishing points detected are based on truly straight parallel lines, improving overall system precision.
3Measurement precision
If vanishing point tracking with prediction filter is implemented, then image stabilization precision is improved, but loss of time increases due to filtering operations
Solution Approach 1:
The patent implements a prediction filter (such as Kalman filter) that uses feedback from previously tracked vanishing point positions to predict current camera motion. This feedback mechanism smooths the motion estimation over time, reducing noise and improving precision while maintaining real-time performance through efficient recursive updating rather than exhaustive processing.
Data Source
AI summary
A method and a device for image stabilization of an image sequence of one or more cameras of a vehicle are disclosed. Images of the surroundings of the vehicle are recorded and image points of the recorded images are projected to a unit sphere, where the projection includes applying a lens distortion correction. For each of the one or more cameras, a vanishing point is calculated using the projected image points on the unit sphere and a motion of the vanishing point on the unit sphere is tracked. The motion of the vanishing point is used to calculate a corrected projection to a ground plane.