Canon kabushiki kaisha (20240177337). SPATIAL CALIBRATION METHOD simplified abstract
Contents
SPATIAL CALIBRATION METHOD
Organization Name
Inventor(s)
GÉRALD Kergourlay of CHEVAIGNE (FR)
ERIC Nassor of THORIGNE-FOUILLARD (FR)
SPATIAL CALIBRATION METHOD - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240177337 titled 'SPATIAL CALIBRATION METHOD
Simplified Explanation
The method described in the abstract is for spatial calibration of a first image against a second image by determining a projection model between the two images and calibrating the first image based on the determined model.
- The method involves obtaining pairs of points of interest in the two images and iteratively determining a projection model by displacing a point of interest, calculating a projection model candidate, and evaluating the reprojection error.
Potential Applications:
- Image processing and analysis
- Augmented reality systems
- Computer vision technology
Problems Solved:
- Accurate alignment of images
- Improving the quality of image processing
- Enhancing the performance of augmented reality applications
Benefits:
- Increased accuracy in spatial calibration
- Improved image matching and alignment
- Enhanced performance of computer vision systems
Potential Commercial Applications:
- Medical imaging
- Surveillance systems
- Robotics and automation
Possible Prior Art: One possible prior art in this field is the use of feature-based methods for image registration and alignment. These methods involve identifying key points in images and using them to establish correspondences between different views.
Unanswered Questions: 1. How does this method compare to existing techniques for spatial calibration of images? 2. Are there any limitations or constraints in the application of this method to real-world scenarios?
Original Abstract Submitted
according to some embodiments of the disclosure, it is provided a method for spatial calibration of a first image against a second image. after obtaining a plurality of pairs of points of interest, each pair matching a point of interest in the first image to a corresponding point of interest in the second image, the method further comprises determining, iteratively, a projection model between the two images and calibrating the first image based on the determined projected model. an iteration of the determining comprises displacing a position of a point of interest, determining a projection model candidate using multiple pairs of points of interest and evaluating a reprojection error between the positions of the points of interest as projected using the projection model candidate and the corresponding points of interest.
