Sub-pixel registration and splicing technology based on interpolation and iterative optimization algorithm
An iterative optimization and image stitching technology, applied in the field of computer vision, can solve the problems of low accuracy of transformation matrix, time-consuming traditional algorithms, occupation of computing resources, etc., to achieve the effect of improving accuracy and stitching effect
Inactive Publication Date: 2019-07-02
BEIJING INSTITUTE OF TECHNOLOGYGY +1
View PDF2 Cites 2 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0008] The purpose of the present invention is to solve the problems of time-consuming traditional algorithms, occupying computing resources, low accuracy of feature matching, and low accuracy of transformation matrix in the process of image stitching in the current image registration process. Pixel registration and splicing technology, including: 1. Artificially setting feature areas to extract feature points and rough matching; 2. Using bilinear interpolation algorithm to interpolate feature matching points to obtain sub-pixel level feature matching points; 3. Using iterative algorithm to refine the image transformation matrix for image stitching
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0019] 1. Sub-pixel registration technology based on bilinear interpolation
[0020] Fix the high-definition camera on a tripod, and use the camera to take two images of the same scene from different angles of view (there is a partial overlapping area between the images), and take the first one as the original image, such as figure 1 , the second one is the image to be registered, such as figure 2 . Artificially set the feature area in the two images to extract and roughly match the feature points (note: the feature points should be spaced as far as possible); perform bilinear interpolation on the pixel-level feature matching points, accurate to the sub-pixel level, use Subpixel-level feature matching points for image registration.
[0021] The registration technology involves the mean square error method and bilinear interpolation algorithm. The registration technology flow chart is as follows: image 3 shown.
[0022] Specific steps:
[0023] When using a high-definiti...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
When an engineering project is carried out, a large-view-angle image can be formed by utilizing an image splicing technology and is further used for engineering building detection, and a shot photo with an overlapped part can be quickly and accurately spliced by utilizing a sub-pixel registration and splicing technology based on an interpolation and iterative optimization algorithm. According to the invnetion, a camera is used for shooting two pictures at different visual angles (with a certain overlap). The method comprises the following steps: taking a first image as an original image and asecond image as a to-be-registered image, firstly preprocessing a photo, respectively finding out a maximum angular point from four characteristic regions set in the to-be-registered image as characteristic points, artificially finding out four corresponding characteristic points in the original image, and simultaneously setting four characteristic regions by taking the characteristic points as centers to carry out characteristic coarse matching; performing sub-pixel registration by using a bilinear interpolation and mean square error algorithm; and then refining the transformation matrix by using an iterative optimization algorithm, and finally realizing image splicing by using the transformation matrix to ensure the rapid and high-precision splicing requirement.
Description
technical field [0001] The invention relates to a sub-pixel registration and mosaic technology, in particular to a sub-pixel registration and mosaic technology based on interpolation and iterative optimization algorithms, belonging to the fields of computer vision and image processing. Background technique [0002] Image registration is the process of matching and superimposing two or more images acquired at different times and under different conditions. It is widely used in the fields of remote sensing data analysis, computer vision, and image processing. Image registration is the basis of image stitching and an important part of image stitching. High-precision image registration technology can ensure a good stitching effect. The main steps of image registration and stitching are feature extraction, feature matching, image registration, image transformation, and image stitching. [0003] Current image registration and stitching techniques usually have the following proble...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06T3/00G06T5/50
CPCG06T5/50G06T3/14
Inventor 许向阳董林石晶欣
Owner BEIJING INSTITUTE OF TECHNOLOGYGY