Description: 摘要:提出了一种基于兴趣点不变矩(IPIM)的图像拼接技术。利用Harris 角检测器获取图像中的兴趣点,计算
兴趣点邻域的平移、旋转及尺度不变矩,通过比较各兴趣点邻域不变矩的欧式距离提取出初始特征点对,根据
几何变换模型剔除伪特征对,最后利用正确映射模型实现图像的拼接。实验表明,该方法对平移以及任意角度
的旋转具有良好的鲁棒性,对于具有小尺度变换的图像仍然具有很好的拼接效果。-Abstract: This paper presents a point of interest based on moment invariants (IPIM) of the image stitching technology. Angle using Harris detector to obtain images in the interest of points to calculate the interest point neighborhood of the translation, rotation and scale invariant moments, by comparing the interest point neighborhood of the Continental moment invariants extracted from the initial feature points right, according to the geometric pseudo-transformation model features removed, and finally realize the use of mapping model correctly spliced images. Experiments show that the method of translation and rotation of any angle has good robustness for small-scale transformation of the image still has a good effect splicing. Platform: |
Size: 323584 |
Author:xuyuhua |
Hits:
Description: In this paper, a novel scale- and rotation-invariant interest point detector and descriptor, coined SURF (Speeded Up Robust
Features) is presented. It approximates or even outperforms previously proposed
schemes with respect to repeatability, distinctiveness, and robustness, yet
can be computed and compared much faster.
This is achieved by relying on integral images for image convolutions by building on the strengths of the leading existing detectors and descriptors (in casu, using a Hessian matrix-based measure for the detector, and a
distribution-based descriptor) and by simplifying these methods to the
essential. This leads to a combination of novel detection, description, and
matching steps. The paper presents experimental results on a standard
evaluation set, as well as on imagery obtained in the context of a real-life
object recognition application. Both show SURF’s strong performance.-In this paper, a novel scale- and rotation-invariant interest point detector and descriptor, coined SURF (Speeded Up Robust
Features) is presented. It approximates or even outperforms previously proposed
schemes with respect to repeatability, distinctiveness, and robustness, yet
can be computed and compared much faster.
This is achieved by relying on integral images for image convolutions by building on the strengths of the leading existing detectors and descriptors (in casu, using a Hessian matrix-based measure for the detector, and a
distribution-based descriptor) and by simplifying these methods to the
essential. This leads to a combination of novel detection, description, and
matching steps. The paper presents experimental results on a standard
evaluation set, as well as on imagery obtained in the context of a real-life
object recognition application. Both show SURF’s strong performance. Platform: |
Size: 686080 |
Author:yangwei |
Hits:
Description: 整合了模式识别领域中几种经典的关于区域识别的文章，对于研究模式识别的同学，会有很大帮助-Integrated field of pattern recognition several classic articles on regional recognition, pattern recognition for research students, there will be a great help Platform: |
Size: 5568512 |
Author:房英明 |
Hits:
Description: C++ implementation of performant scale- and rotation-invariant interest point detector and descriptor SURF: Speeded Up Robust Features Platform: |
Size: 104448 |
Author:Athos |
Hits:
Description: 仿射不变Harris, Laplacian, det(Hessian) and Ridge 特征点检测
参考文献：An affine invariant interest point detector , K.Mikolajczyk and C.Schmid, ECCV 02, pp.I:128-142.-Matlab code for detecting Affine spatial interest points. Includes Harris, Laplacian, det(Hessian) and Ridge interest point operators in combination with spatial scale selection based on the extrema of scale-normalized Laplacian and affine adaptation basen on second-moment matrix. Scale and shape adaptation are optional and disjoint.
Zip archive: affintpoints.zip
Ref: An affine invariant interest point detector , K.Mikolajczyk and C.Schmid, ECCV 02, pp.I:128-142.
What is in the package:
1) ineterst point detection of different kinds (Harris, Laplace, det(H), Ridge)
2) scale, shape and position adaptation procedure
3) demo examples and a script for batch-mode computation and saving of the results
what is not in the package:
- no rotation estimation
- no region descriptor computation Platform: |
Size: 901120 |
Author:灵台斜月 |
Hits:
Search - Scale Invariant Interest Point Detector