A gray-scale template matching algorithm with rotation and zoom invariance
A technology of template matching and algorithm, applied in still image data retrieval, still image data query, etc., can solve the problems of single output information, loss of matching accuracy, slow speed, etc., and achieve good matching stability, high matching accuracy, shortened matching the effect of time
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0036] Embodiment 1: Compared with the template, the picture to be matched has no scaling and no rotation.
[0037] as attached Figure 1-8 As shown, a gray-scale template matching algorithm with rotation and scaling invariance, the algorithm operation process includes: ring matching, radial line matching and discrete matching; specific matching steps:
[0038] 1) Ring matching: Before creating a template, you need to set the zoom matching parameters, MinScale: 0.3, MaxScale: 1.0, which means that the template size is used as the benchmark, and the zoom search is performed by 0.3-1.0 times; first, the center point of the template image is used as the center. The largest inscribed square, and scale the square according to the set scaling parameters, search interval: 0.1, take the gray value of the pixel on n rings in each square picture, n=square side length / 2-1, circle The center of the ring is the center of the square, and the average gray value t1, t2..., tn of the pixels o...
Embodiment 2
[0042] Embodiment 2: The picture to be matched is reduced by 50% compared with the template, and there is no rotation.
[0043] as attached Figure 1-4 As shown in and 9-12, a gray-scale template matching algorithm with rotation and scaling invariance, the algorithm operation process includes: ring matching, radial line matching and discrete matching; specific matching steps:
[0044] 1) Ring matching: Before creating a template, you need to set the zoom matching parameters, MinScale: 0.3, MaxScale: 1.0, which means that the template size is used as the benchmark, and the zoom search is performed by 0.3-1.0 times; first, the center point of the template image is used as the center. The largest inscribed square, and scale the square according to the set scaling parameters, search interval: 0.1, take the gray value of the pixel on n rings in each square picture, n=square side length / 2-1, circle The center of the ring is the center of the square, and the average gray value t1, t...
Embodiment 3
[0048] Embodiment 3: The picture to be matched and the template are not scaled, and rotated 35° counterclockwise.
[0049] as attached Figure 1-4 As shown in and 13-16, a gray-scale template matching algorithm with rotation and scaling invariance, the algorithm operation process includes: ring matching, radial line matching and discrete matching; specific matching steps:
[0050] 1) Ring matching: Before creating a template, you need to set the zoom matching parameters, MinScale: 0.3, MaxScale: 1.0, which means that the template size is used as the benchmark, and the zoom search is performed by 0.3-1.0 times; first, the center point of the template image is used as the center. The largest inscribed square, and scale the square according to the set scaling parameters, search interval: 0.1, take the gray value of the pixel on n rings in each square picture, n=square side length / 2-1, circle The center of the ring is the center of the square, and the average gray value t1, t2......
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


