ImageCorrespondingPoints[image1,image2]
找到 image1 和 image2 中匹配的一组兴趣点,并且返回它们的像素坐标.
ImageCorrespondingPoints
ImageCorrespondingPoints[image1,image2]
找到 image1 和 image2 中匹配的一组兴趣点,并且返回它们的像素坐标.
更多信息和选项
- ImageCorrespondingPoints 使用 ImageKeypoints 来寻找候选的对应点.
- ImageCorrespondingPoints[image1,image2] 返回形如 {points1,points2} 的表达式,其中 pointsi 是表示 imagei 中匹配点的像素坐标列表.
- 可以指定下列选项:
-
KeypointStrength Automatic 关键点的最小强度 Masking All 感兴趣的区域 MaxFeatures Automatic 关键点的最大数目 Method Automatic 使用的关键点的类型 TransformationClass None 点之间的几何关系 - 在设置 Masking->roi 下,点集受到限制,使得 image1 返回的点 points1 全部落在感兴趣的区域内.
- 在设置 MaxFeatures->n 下,最多返回具有最大平均关键点强度的 n 个对应点.
- 缺省情况下,使用一个合适的关键点类型来找到对应的点. 通过 Methodmethod,可以指定使用一种特殊的关键点类型或类型列表.
- method 的可能的的设置包括:
-
"AKAZE" 加速 KAZE 和二进制描述符 "BRISK" 二进制鲁棒不变可扩展关键点 ((BRISK) "KAZE" 非线性尺度空间检测器和描述符 "ORB" FAST 检测器和二进制稳健独立基本特征 (BRIEF) 描述符 "SIFT" 尺度不变特征变换 (SIFT) 检测器和描述符 "RootSIFT" 有加强描述符的 SIFT 关键点 "SURF" 快速鲁棒特征 {method1,method2,…} 各种关键点对应关系的组合 - TransformationClass 的可能设置包括:
-
None 没有几何约束 "Translation" 只有平移 "Rigid" 平移和旋转 "Similarity" 平移、旋转、缩放 "Affine" 线性变换和平移 "Perspective" 线性分式变换 "Epipolar" 对极变换,把图像的一个点映射成另一图像的一条线
范例
打开所有单元 关闭所有单元范围 (3)
选项 (16)
KeypointStrength (3)
ImageCorrespondingPoints[[image], [image]]ImageCorrespondingPoints[[image], [image], KeypointStrength -> .001]{i1, i2} = {[image], [image]};ListLinePlot[Table[{s, Length[ImageCorrespondingPoints[i1, i2, KeypointStrength -> s][[1]]]}, {s, .0001, .005, .0001}], Frame -> True, FrameLabel -> {"threshold", "number of corresponding points"}]Masking (1)
默认情况下,设置 Masking->All 返回所有探测到的对应点:
{i1, i2} = {[image], [image]};MapThread[HighlightImage[#1, #2]&, {{i1, i2}, ImageCorrespondingPoints[i1, i2]}]设置 Masking->maskimage,image1 的相应点位于 maskimage 内:
mask = [image];
c = ImageCorrespondingPoints[i1, i2, Masking -> mask];HighlightImage[i1, {c[[1]], Orange, mask}]MaxFeatures (2)
c = ImageCorrespondingPoints[[image], [image], MaxFeatures -> 5]返回的对应点数可能少于 MaxFeatures 选项的数值:
ImageCorrespondingPoints[[image], [image], KeypointStrength -> .001, MaxFeatures -> 5]//First//LengthMethod (3)
ImageCorrespondingPoints[[image], [image]]ImageCorrespondingPoints[[image], [image], Method -> "KAZE"]ImageCorrespondingPoints[[image], [image], Method -> {"SURF", "BRISK"}]//ShortTransformationClass (7)
ImageCorrespondingPoints[[image], [image]]//First//LengthImageCorrespondingPoints[[image], [image], TransformationClass -> "Epipolar"]//First//LengthImageCorrespondingPoints[[image], [image], TransformationClass -> "Perspective"]//First//LengthImageCorrespondingPoints[[image], [image], TransformationClass -> "Affine"]//First//LengthImageCorrespondingPoints[[image], [image], TransformationClass -> "Similarity"]//First//LengthImageCorrespondingPoints[[image], [image], TransformationClass -> "Rigid"]//First//LengthImageCorrespondingPoints[[image], [image], TransformationClass -> "Translation"]应用 (3)
ImageCorrespondingPoints[[image], [image], TransformationClass -> "Perspective"]i1 = [image];i2 = [image];
matches = ImageCorrespondingPoints[i1, i2, "Transformation" -> "Perspective"];
MapThread[ImageTrim, {{i1, i2}, matches}]i1 = [image];i2 = [image];
matches = ImageCorrespondingPoints[i1, i2];
a1 = ArcTan@@@(# - 1 / 2 * ImageDimensions@i1& /@ matches[[1]]);
a2 = ArcTan@@@(# - 1 / 2 * ImageDimensions@i2& /@ matches[[2]]);
RootMeanSquare[Mod[a1 - a2, 2 Pi]] / Degree属性和关系 (1)
ImageCorrespondingPoints 把所有图像转换成灰度图:
{i1, i2} = {[image], [image]};ImageCorrespondingPoints[i1, i2] == ImageCorrespondingPoints[Sequence@@ColorConvert[{i1, i2}, "Grayscale"]]文本
Wolfram Research (2010),ImageCorrespondingPoints,Wolfram 语言函数,https://reference.wolfram.com/language/ref/ImageCorrespondingPoints.html (更新于 2021 年).
CMS
Wolfram 语言. 2010. "ImageCorrespondingPoints." Wolfram 语言与系统参考资料中心. Wolfram Research. 最新版本 2021. https://reference.wolfram.com/language/ref/ImageCorrespondingPoints.html.
APA
Wolfram 语言. (2010). ImageCorrespondingPoints. Wolfram 语言与系统参考资料中心. 追溯自 https://reference.wolfram.com/language/ref/ImageCorrespondingPoints.html 年
BibTeX
@misc{reference.wolfram_2026_imagecorrespondingpoints, author="Wolfram Research", title="{ImageCorrespondingPoints}", year="2021", howpublished="\url{https://reference.wolfram.com/language/ref/ImageCorrespondingPoints.html}", note=[Accessed: 17-November-2026]}
BibLaTeX
@online{reference.wolfram_2026_imagecorrespondingpoints, organization={Wolfram Research}, title={ImageCorrespondingPoints}, year={2021}, url={https://reference.wolfram.com/language/ref/ImageCorrespondingPoints.html}, note=[Accessed: 17-November-2026]}