ImageDistances.jl

ImageDistances.jl aims to:

• follow the same API in Distances.jl
• support image types
• provide image-specific distances

Usage

ImageDistances.jl is shipped together with Images.jl, but you can still use it as a standalone package.

# both line includes this package
using Images
using ImageDistances

Here's a simple usage example

using ImageDistances, TestImages

d = Euclidean()
imgA = testimage("cameraman") # N0f8
imgB = testimage("lena_gray_512") # N0f8

# distance between two images
evaluate(d, imgA, imgB) # 142.59576f0
d(imgA, imgB) # 142.59576f0

Distances are calculated regardless of the color type and storage type.

using ImageCore

# For gray image, all of them equals
d(imgA, imgB) # 142.59576f0
d(float32.(imgA), float32.(imgB)) # 142.59576f0
d(Float32.(imgA), Float32.(imgB)) # 142.59576f0
d(imgA, float32.(imgB)) # 142.59576f0

However, for Color3 images such as RGB, it's noteworthy that the following results are different in general.

d = ZNCC()
imgA = testimage("lena_color_512")
imgB = testimage("fabio_color_512")
# distance of each pixel is calculated first, and then sum up all pixels
d(imgA, imgB) # 0.023451565f0
# distance of each slice is calculated first, and then sum up three channels
d(channelview(imgA), channelview(imgB)) # 0.21142173f0

That's said, to achieve the same results to other languages, you need to channelview the image first to get a raw numeric view.

Just like in Distances.jl, huge performance gains are obtained by calling the colwise and pairwise functions instead of naively looping over a collection of images and calling evaluate.

d = ModifiedHausdorff()

# two lists of images
imgsA = [imgA, imgB, ...]
imgsB = [imgB, imgA, ...]

# distance between the "columns"
colwise(d, imgsA, imgsB)

# distance between every pair of images
pairwise(d, imgsA, imgsB)
pairwise(d, imgsA)

Distances

General Distances

type name	convenient syntax	math definition
Euclidean	euclidean(x, y)	sqrt(sum((x - y) .^ 2))
SqEuclidean	sqeuclidean(x, y)	sum((x - y).^2)
Cityblock	cityblock(x, y)	sum(abs(x - y))
TotalVariation	totalvariation(x, y)	sum(abs(x - y)) / 2
Minkowski	minkowski(x, y, p)	sum(abs(x - y).^p) ^ (1/p)
Hamming	hamming(x, y)	sum(x .!= y)
SumAbsoluteDifference	sad(x, y)	sum(abs(x - y))
SumSquaredDifference	ssd(x, y)	sum((x - y).^2)
MeanAbsoluteError	mae(x, y), sadn(x, y)	sum(abs(x - y))/len(x)
MeanSquaredError	mse(x, y), ssdn(x, y)	sum((x - y).^2)/len(x)
RootMeanSquaredError	rmse(x, y)	sqrt(sum((x - y) .^ 2)/len(x))
ZNCC	zncc(x, y)	dot(x,y)/(norm(x)*norm(y))

Image-specific Distances

Distance type	Convenient syntax	References
Hausdorff and ModifiedHausdorff	hausdorff(imgA,imgB) and modified_hausdorff(imgA,imgB)	Dubuisson, M-P et al. 1994. A Modified Hausdorff Distance for Object-Matching.
CIEDE2000	ciede2000(imgA,imgB) and ciede2000(imgA,imgB; metric=DE_2000())	Sharma, G., Wu, W., and Dalal, E. N., 2005. The CIEDE2000 color‐difference formula.