将 Java 中的 3D 仿射变换构建导入 R

Question

我有一个与仿射变换（<affine>标签之间）相关的问题。我正在使用AffineTransform3D函数从使用 Java 创建的 BigDataViewer Fiji 插件创建的 xml 文件（此处为完整 xml 文件）中提取以下两个仿射转换：

<ViewRegistrations>

<ViewRegistration timepoint="0" setup="0">

  <ViewTransform type="affine">

    <Name>Fast 3d geometric hashing (rotation invariant), AffineModel3D on [beads (c=0)]</Name>

    <affine>1.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 1.0 0.0</affine>

  </ViewTransform>

  <ViewTransform type="affine">

    <Name>calibration</Name>

    <affine>1.9998662296836334 0.0 0.0 0.0 0.0 1.9998662296836334 0.0 0.0 0.0 0.0 1.9998662296836334 0.0</affine>

  </ViewTransform>

</ViewRegistration>

我想在 R 中导入两个仿射变换，使用buildAffine()R 包 {RNiftyReg} 中的函数，然后使用composeTransforms(){RNiftyReg}计算它们的组成。

buildAffine(translation = c(0, 0, 0), scales = c(1, 1, 1), skews = c(0, 0,0),

  angles = c(0, 0, 0), source = NULL, target = NULL,

  anchor = c("none", "origin", "centre", "center"))

我的问题：

上面的仿射变换存储在一个包含 12 个索引的向量中。buildAffine()需要将平移、比例、倾斜和角度的值作为输入参数。

我想知道哪个值对应什么。

Answer 1

我主要是 R 用户，但这里是： hte Java 调用中的变量名称是：&nbsp;dat <- scan (text="double xx, double yx, double zx, double tx, double xy, double yy, double zy, double ty, double xz, double yz, double zz, double tz", what="")dat <- dat[dat != "double"]matrix(dat,4)&nbsp; &nbsp; &nbsp; [,1]&nbsp; [,2]&nbsp; [,3]&nbsp;[1,] "xx," "xy," "xz,"[2,] "yx," "yy," "yz,"[3,] "zx," "zy," "zz,"[4,] "tx," "ty," "tz"&nbsp;此页面记录了如何对仿射变换进行编码，唯一的区别是转置了 4x3 矩阵：https : //o7planning.org/en/11157/javafx-transformations-tutorial&nbsp;t( matrix(dat,4) )&nbsp; &nbsp; &nbsp;[,1]&nbsp; [,2]&nbsp; [,3]&nbsp; [,4]&nbsp;[1,] "xx," "yx," "zx," "tx,"[2,] "xy," "yy," "zy," "ty,"[3,] "xz," "yz," "zz," "tz"&nbsp;所以矩阵将应用于 c(x,y,z,1) 的向量以获得输出：