Warp Affine using R

The getAffineTransform() and warpAffine() functions of the OpenImageR package is an RcppArmadillo re-implementation of existing Python Code and this vignette shows how these functions can be used from within R based on the author's .ipynb file


require(OpenImageR)

path = system.file('tmp_images', 'landscape.jpg', package = "OpenImageR")

img = readImage(path)
print(dim(img))


knitr::include_graphics(path)


r = ncol(img)
c = nrow(img)
offset = 50

original_points = matrix(data = c(0, 0, r, 0, 0, c),
                         nrow = 3,
                         ncol = 2,
                         byrow = TRUE)

transformed_points = matrix(data = c(offset, 0, r, offset, 0, c-offset),
                            nrow = 3,
                            ncol = 2,
                            byrow = TRUE)

M_aff = getAffineTransform(original_points = original_points,
                           transformed_points = transformed_points)


The following is the Affine transformation matrix,

print(M_aff)


The Affine transformation matrix can be used as input in the warpAffine() function,

res_3d = warpAffine(img = img,
                    M = M_aff,
                    R = r,
                    C = c,
                    verbose = TRUE)

str(res_3d)


The next image shows the output based on the input data and parameters,

imageShow(res_3d, clear_viewer = FALSE)




Try the OpenImageR package in your browser

Any scripts or data that you put into this service are public.

OpenImageR documentation built on Aug. 24, 2022, 5:07 p.m.