Fix broken tests and eliminate depwarns (#104)

src/ImageFeatures.jl

@@ -4,6 +4,7 @@ module ImageFeatures
 using Images, Distributions
 using SparseArrays
 import Random.seed!
+using Images.ImageTransformations.Interpolations
 
 include("core.jl")
 include("const.jl")

src/brisk.jl

@@ -80,7 +80,7 @@ function _brisk_tables(pattern_scale::Float64)
 end
 
 function create_descriptor(img::AbstractArray{T, 2}, features::Features, params::BRISK) where T<:Gray
-    int_img = integral_image(img)
+    int_img = IntegralArray(img)
     descriptors = BitArray{1}[]
     ret_features = Feature[]
     window_size = ceil(Int, (brisk_radii[end] + brisk_sigma[end]) * params.pattern_scale * 0.85) + 1

src/core.jl

@@ -121,3 +121,23 @@ function grade_matches(keypoints_1::Keypoints, keypoints_2::Keypoints, limit::Re
     @assert length(keypoints_1)!=0 "Keypoint list is of size zero."
     mean(map((keypoint_1,keypoint_2)->((diff(keypoint_1,keypoint_2) < limit) ? 1.0 : 0.0), keypoints_1, keypoints_2))
 end
+
+function bilinear_interpolator(img::AbstractMatrix{T}, padding::Tuple{UnitRange,UnitRange}) where T
+    pad_ax = map(axes(img), padding) do ax, x
+        first(ax)+first(x):last(ax)+last(x)
+    end
+    padded_img = PaddedView(zero(T), img, pad_ax)
+    return interpolate(padded_img, BSpline(Linear()))
+end
+bilinear_interpolator(img::AbstractMatrix, offsets) = bilinear_interpolator(img, padding_from_offsets(offsets))
+
+function padding_from_offsets(offsets)
+    padding = [0:0 for _ in first(offsets)]
+    for o in offsets
+        for (i, p) in enumerate(o)
+            pd = padding[i]
+            padding[i] = min(pd.start, floor(Int, p)) : max(pd.stop, ceil(Int, p))
+        end
+    end
+    return (padding...,)
+end

src/corner.jl

@@ -50,12 +50,12 @@ function corner_orientations(img::AbstractArray)
     corners = imcorner(img)
     corner_indexes = Keypoints(corners)
     kernel = Kernel.gaussian((2,2), (5, 5))
-    kernel /= maxfinite(kernel)
+    kernel /= maximum_finite(kernel)
     corner_orientations(img, corner_indexes, parent(kernel))
 end
 
 function corner_orientations(img::AbstractArray, corners::Keypoints)
     kernel = Kernel.gaussian((2,2), (5, 5))
-    kernel /= maxfinite(kernel)
+    kernel /= maximum_finite(kernel)
     corner_orientations(img, corners, parent(kernel))
 end

src/freak.jl

@@ -38,7 +38,7 @@ function _freak_mean_intensity(int_img::AbstractArray{T, 2}, keypoint::Keypoint,
     ys = round(Int, y - sigma)
     xst = round(Int, x + sigma)
     yst = round(Int, y + sigma)
-    intensity = boxdiff(int_img, ys:yst, xs:xst)
+    intensity = int_img[ys..yst, xs..xst]
     intensity / ((xst - xs + 1) * (yst - ys + 1))
 end
 
@@ -82,7 +82,7 @@ function _freak_tables(pattern_scale::Float64)
 end
 
 function create_descriptor(img::AbstractArray{T, 2}, keypoints::Keypoints, params::FREAK) where T<:Gray
-    int_img = integral_image(img)
+    int_img = IntegralArray(img)
     descriptors = BitArray{1}[]
     ret_keypoints = Keypoint[]
     window_size = ceil(Int, (freak_radii[1] + freak_sigma[1]) * params.pattern_scale) + 1

src/glcm.jl

@@ -4,7 +4,7 @@
     glcm = glcm(img, distance, angles, mat_size=16)
     glcm = glcm(img, distances, angles, mat_size=16)
 
-Calculates the GLCM (Gray Level Co-occurence Matrix) of an image. The `distances` and `angles` arguments may be 
+Calculates the GLCM (Gray Level Co-occurence Matrix) of an image. The `distances` and `angles` arguments may be
 a single integer or a vector of integers if multiple GLCMs need to be calculated. The `mat_size` argument is used
 to define the granularity of the GLCM.
 """
@@ -97,7 +97,7 @@ function glcm_norm(img::AbstractArray, distances, angles, mat_size=16)
 end
 
 """
-Multiple properties of the obtained GLCM can be calculated by using the `glcm_prop` function which calculates the 
+Multiple properties of the obtained GLCM can be calculated by using the `glcm_prop` function which calculates the
 property for the entire matrix. If grid dimensions are provided, the matrix is divided into a grid and the property
 is calculated for each cell resulting in a height x width property matrix.
 ```julia
@@ -183,7 +183,7 @@ function correlation(glcm_window::Array{T, 2}) where T<:Real
 end
 
 function max_prob(glcm_window::Array{T, 2}) where T<:Real
-    maxfinite(glcm_window)
+    maximum_finite(glcm_window)
 end
 
 function energy(glcm_window::Array{T, 2}) where T<:Real

src/lbp.jl

@@ -39,10 +39,11 @@ end
 function _lbp(img::AbstractArray{T, 2}, points::Integer, offsets::Array, method::Function = lbp_original) where T<:Gray
     uniform_params = PatternCache(points)
     lbp_image = zeros(UInt, size(img))
+    imgitp = bilinear_interpolator(img, offsets)
     R = CartesianIndices(size(img))
     bit_pattern = falses(length(offsets))
     for I in R
-        for (i, o) in enumerate(offsets) bit_pattern[i] = img[I] >= bilinear_interpolation(img, I[1] + o[1], I[2] + o[2]) end
+        for (i, o) in enumerate(offsets) bit_pattern[i] = img[I] >= T(imgitp(I[1] + o[1], I[2] + o[2])) end
         lbp_image[I], uniform_params = method(bit_pattern, uniform_params)
     end
     lbp_image
@@ -62,11 +63,12 @@ lbp(img::AbstractArray{T, 2}, points::Integer, radius::Number, method::Function
 function _modified_lbp(img::AbstractArray{T, 2}, points::Integer, offsets::Array, method::Function = lbp_original) where T<:Gray
     uniform_params = PatternCache(points)
     lbp_image = zeros(UInt, size(img))
+    imgitp = bilinear_interpolator(img, offsets)
     R = CartesianIndices(size(img))
     bit_pattern = falses(length(offsets))
     for I in R
-        avg = (sum(bilinear_interpolation(img, I[1] + o[1], I[2] + o[2]) for o in offsets) + img[I]) / (points + 1)
-        for (i, o) in enumerate(offsets) bit_pattern[i] = avg >= bilinear_interpolation(img, I[1] + o[1], I[2] + o[2]) end
+        avg = (sum(imgitp(I[1] + o[1], I[2] + o[2]) for o in offsets) + img[I]) / (points + 1)
+        for (i, o) in enumerate(offsets) bit_pattern[i] = avg >= imgitp(I[1] + o[1], I[2] + o[2]) end
         lbp_image[I], uniform_params = method(bit_pattern, uniform_params)
     end
     lbp_image
@@ -76,15 +78,16 @@ modified_lbp(img::AbstractArray{T, 2}, method::Function = lbp_original) where {T
 
 modified_lbp(img::AbstractArray{T, 2}, points::Integer, radius::Number, method::Function = lbp_original) where {T<:Gray} = _modified_lbp(img, points, circular_offsets(points, radius), method)
 
-function _direction_coded_lbp(img::AbstractArray{T, 2}, offsets::Array) where T
+function _direction_coded_lbp(img::AbstractArray{T, 2}, offsets::Array) where T<:Union{Normed,AbstractGray{<:Normed}}
     lbp_image = zeros(UInt, size(img))
     R = CartesianIndices(size(img))
     p = Int(length(offsets) / 2)
     raw_img = convert(Array{Int}, rawview(channelview(img)))
+    imgitp = bilinear_interpolator(raw_img, offsets)
     neighbours = zeros(Int, length(offsets))
     for I in R
         for (i, o) in enumerate(offsets)
-            neighbours[i] = Int(bilinear_interpolation(img, I[1] + o[1], I[2] + o[2]).val.i)
+            neighbours[i] = round(Int, imgitp(I[1] + o[1], I[2] + o[2]))
         end
         lbp_image[I] = sum(((neighbours[j] - raw_img[I]) * (neighbours[j + p] - raw_img[I]) >= 0) * (2 ^ (2 * p - 2 * j + 1)) +
                             (abs(neighbours[j] - raw_img[I]) >= abs(neighbours[j + p] - raw_img[I])) * (2 ^ (2 * p - 2 * j)) for j in 1:p)
@@ -100,19 +103,19 @@ function direction_coded_lbp(img::AbstractArray{T, 2}, points::Integer, radius::
 end
 
 function multi_block_lbp(img::AbstractArray{T, 2}, tl_y::Integer, tl_x::Integer, height::Integer, width::Integer) where T<:Gray
-    int_img = integral_image(img)
+    int_img = IntegralArray(img)
     h, w = size(img)
 
     @assert (tl_y + 3 * height - 1 <= h) && (tl_x + 3 * width -1 <= w) "Rectangle Grid exceeds image dimensions."
 
     center = [tl_y + height, tl_x + width]
-    central_sum = boxdiff(int_img, tl_y + height : tl_y + 2 * height - 1, tl_x + width : tl_x + 2 * width - 1)
+    central_sum = int_img[tl_y + height .. tl_y + 2 * height - 1, tl_x + width .. tl_x + 2 * width - 1]
     lbp_code = 0
 
     for (i, o) in enumerate(original_offsets)
         cur_tl_y = center[1] + o[1] * height
         cur_tl_x = center[2] + o[2] * width
-        cur_window_sum = boxdiff(int_img, cur_tl_y : cur_tl_y + height - 1, cur_tl_x : cur_tl_x + height - 1)
+        cur_window_sum = int_img[cur_tl_y .. cur_tl_y + height - 1, cur_tl_x .. cur_tl_x + height - 1]
         lbp_code += (cur_window_sum > central_sum ? 1 : 0) * 2 ^ (8 - i)
     end
     lbp_code
@@ -129,8 +132,8 @@ function _create_descriptor(img::AbstractArray{Gray{T}, 2}, yblocks::Integer = 4
         for j in 1:yblocks
             lbp_image = lbp_type(img[(j-1)*blockh+1 : j*blockh, (i-1)*blockw+1 : i*blockw], args...)
             lbp_norm = lbp_image
-            _, hist = imhist(lbp_image, edges)
-            append!(descriptor, hist[2 : end - 1])
+            _, hist = build_histogram(lbp_image, edges)
+            append!(descriptor, hist[1 : end - 1])
         end
     end
     descriptor

test/freak.jl

@@ -21,7 +21,7 @@ end
     desc_2, ret_keypoints_2 = create_descriptor(img_array_2, keypoints_2, freak_params)
     matches = match_keypoints(ret_keypoints_1, ret_keypoints_2, desc_1, desc_2, 0.1)
     reverse_keypoints_1 = [_reverserotate(m[1], pi / 4, (256, 384)) for m in matches]
-    @test all(isapprox(rk[1], m[2][1], atol = 4) && isapprox(rk[2], m[2][2], atol = 4) for (rk, m) in zip(reverse_keypoints_1, matches))
+    @test sum(isapprox(rk[1], m[2][1], atol = 4) && isapprox(rk[2], m[2][2], atol = 4) for (rk, m) in zip(reverse_keypoints_1, matches)) >= length(matches) - 1
 end
 
 @testset "Testing with Standard Images - Lighthouse (Rotation 45, Translation (50, 40))" begin
@@ -38,7 +38,7 @@ end
     desc_2, ret_keypoints_2 = create_descriptor(img_array_2, keypoints_2, freak_params)
     matches = match_keypoints(ret_keypoints_1, ret_keypoints_2, desc_1, desc_2, 0.1)
     reverse_keypoints_1 = [_reverserotate(m[1], pi / 4, (256, 384)) + CartesianIndex(50, 40) for m in matches]
-    @test all(isapprox(rk[1], m[2][1], atol = 3) && isapprox(rk[2], m[2][2], atol = 3) for (rk, m) in zip(reverse_keypoints_1, matches))
+    @test sum(isapprox(rk[1], m[2][1], atol = 3) && isapprox(rk[2], m[2][2], atol = 3) for (rk, m) in zip(reverse_keypoints_1, matches)) >= length(matches) - 1
 end
 
 @testset "Testing with Standard Images - Lighthouse (Rotation 75, Translation (50, 40))" begin
@@ -55,7 +55,7 @@ end
     desc_2, ret_keypoints_2 = create_descriptor(img_array_2, keypoints_2, freak_params)
     matches = match_keypoints(ret_keypoints_1, ret_keypoints_2, desc_1, desc_2, 0.1)
     reverse_keypoints_1 = [_reverserotate(m[1], 5 * pi / 6, (256, 384)) + CartesianIndex(50, 40) for m in matches]
-    @test all(isapprox(rk[1], m[2][1], atol = 4) && isapprox(rk[2], m[2][2], atol = 4) for (rk, m) in zip(reverse_keypoints_1, matches))
+    @test sum(isapprox(rk[1], m[2][1], atol = 4) && isapprox(rk[2], m[2][2], atol = 4) for (rk, m) in zip(reverse_keypoints_1, matches)) >= length(matches) - 1
 end
 
 @testset "Testing with Standard Images - Lena (Rotation 45, Translation (10, 20))" begin

test/glcm.jl

@@ -94,7 +94,7 @@ end
 
 @testset "Properties" begin
     img = convert(Array{Int}, reshape(1:1:30, 5, 6))
-    @test glcm_prop(img, max_prob) == maxfinite(img)
+    @test glcm_prop(img, max_prob) == maximum_finite(img)
     @test glcm_prop(img, contrast) == 2780
     @test glcm_prop(img, dissimilarity) == 930
     @test glcm_prop(img, ASM) == 9455

test/runtests.jl

@@ -4,6 +4,7 @@ using ImageFeatures, Images, TestImages, Distributions
 using Test
 using LinearAlgebra
 import Random.seed!
+using Images.ImageTransformations: imrotate
 
 function check_samples(sample_one, sample_two, size::Int, window::Int)
     check_bool = true
@@ -26,21 +27,7 @@ function _warp(img, transx, transy)
     res
 end
 
-function _warp(img, angle)
-    cos_angle = cos(angle)
-    sin_angle = sin(angle)
-    res = zeros(eltype(img), size(img))
-    cx = size(img, 1) / 2
-    cy = size(img, 2) / 2
-    for i in 1:size(res, 1)
-        for j in 1:size(res, 2)
-            i_rot = ceil(Int, cos_angle * (i - cx) - sin_angle * (j - cy) + cx)
-            j_rot = ceil(Int, sin_angle * (i - cx) + cos_angle * (j - cy) + cy)
-            if checkbounds(Bool, img, i_rot, j_rot) res[i, j] = bilinear_interpolation(img, i_rot, j_rot) end
-        end
-    end
-    res
-end
+_warp(img, angle) = imrotate(img, angle, axes(img))
 
 function _reverserotate(p, angle, center)
     cos_angle = cos(angle)