Stereo Depth Calculation Using Fisheye and Projective Cameras
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle driving assistance systems face challenges in accurately detecting objects and calculating distances using cameras with different lens types and positions, leading to distortions and reduced accuracy in object detection and distance calculation.
Innovation Solution
The system employs image data from cameras with different lens types, such as fisheye and projective lenses, to determine a point in space by matching pixels and calculating distances based on the lens types and positional offsets, allowing for accurate object detection and distance calculation despite camera differences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If cameras with different lens types (fisheye and projective) are used to provide comprehensive coverage, then the field of view and coverage area are improved, but the measurement precision and accuracy of object detection and distance calculation deteriorate due to distortions
Solution Approach 1:
The system changes the parameters of image processing by applying different distortion correction models and calibration methods for fisheye and projective lenses. By adjusting these parameters, the system maintains comprehensive coverage while correcting the measurement inaccuracies introduced by different lens types.
Solution Approach 2:
The system introduces an intermediary processing layer that includes a processor configured to receive image data from both lens types, determine camera characteristics, and calculate corrected distances. This intermediary processing mediates between the raw image data from different lenses and the final accurate measurements.
2Adaptability or versatility
If multiple cameras with different characteristics are deployed to enhance situational awareness, then the adaptability and coverage are improved, but the device complexity increases
Solution Approach 1:
The system implements universality by creating a single processing architecture that handles multiple camera types (fisheye, projective, and other lenses) through a unified processor. The processor is configured to accommodate different camera characteristics using generalizable algorithms, allowing one system to handle diverse camera configurations without requiring separate processing paths for each camera type.
3Area of stationary object
If cameras are positioned at different locations and orientations to provide 360 degree view, then the coverage area is improved, but the measurement precision of baseline point determination and distance calculation deteriorates
Solution Approach 1:
The system performs preliminary action by pre-determining camera characteristics including intrinsics and extrinsics for each camera before actual object detection. The processor is configured to receive these pre-calibrated parameters and use them to accurately determine baseline points and calculate distances, accounting for the specific positioning and orientation of each camera in the 360-degree arrangement.
Data Source
AI summary
This disclosure provides systems, methods, and devices for vehicle driving assistance systems that support image processing. In a first aspect, a method of image processing includes receiving first and second image data from first and second cameras having different lens types. A first field of view of the second image data overlaps at least a portion of a second field of view of the first image data. The method further includes determining a point in space based on the first image data and the second image data and calculating a distance between the first camera and the point in space based on the lens type of the first camera and the lens type of the second camera. Other aspects and features are also claimed and described.