SKIMAGE Archives - icodebroker

[PYTHON/SKIMAGE] 수퍼 픽셀(Superpixel) 구하기

■ 수퍼 픽셀(Superpixel)을 구하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
import matplotlib.pyplot as pp
import skimage.color as color
import skimage.io as io
import skimage.segmentation as segmentation

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

imageSegmentNDArray = segmentation.slic(imageNDArray, compactness = 20, n_segments = 500)

superpixelImageNDArray = color.label2rgb(imageSegmentNDArray, imageNDArray, kind="avg")

figure, axesNDArray = pp.subplots(nrows = 1, ncols = 2, sharex = True, sharey = True, figsize = (6, 8))

axesNDArray[0].imshow(imageNDArray)
axesNDArray[1].imshow(superpixelImageNDArray)

for axes in axesNDArray:
    axes.axis("off")

pp.tight_layout()

pp.show()

import matplotlib

import matplotlib.pyplot as pp

import skimage.color as color

import skimage.io as io

import skimage.segmentation as segmentation

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

imageSegmentNDArray = segmentation.slic(imageNDArray, compactness = 20, n_segments = 500)

superpixelImageNDArray = color.label2rgb(imageSegmentNDArray, imageNDArray, kind="avg")

figure, axesNDArray = pp.subplots(nrows = 1, ncols = 2, sharex = True, sharey = True, figsize = (6, 8))

axesNDArray[0].imshow(imageNDArray)

axesNDArray[1].imshow(superpixelImageNDArray)

for axes in axesNDArray:

axes.axis("off")

pp.tight_layout()

pp.show()

source.png

[PYTHON/SKIMAGE] find_contours 함수 : 외곽선 검출하기

■ find_contours 함수를 사용해 외곽선을 검출하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
import matplotlib.pyplot as pp
import skimage.measure as measure
import skimage.io as io
import skimage.color as color
import skimage.filters as filters

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

sobelImageNDArray = filters.sobel(grayscaleImageNDArray)

contourList = measure.find_contours(sobelImageNDArray, 0.2)

figure, axes = pp.subplots()

axes.imshow(sobelImageNDArray, interpolation = "nearest", cmap = pp.cm.gray)

for n, contour in enumerate(contourList):
    axes.plot(contour[:, 1], contour[:, 0], linewidth = 2)

axes.axis("image")

axes.set_xticks([])
axes.set_yticks([])

pp.show()

import matplotlib

import matplotlib.pyplot as pp

import skimage.measure as measure

import skimage.io as io

import skimage.color as color

import skimage.filters as filters

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

sobelImageNDArray = filters.sobel(grayscaleImageNDArray)

contourList = measure.find_contours(sobelImageNDArray, 0.2)

figure, axes = pp.subplots()

axes.imshow(sobelImageNDArray, interpolation = "nearest", cmap = pp.cm.gray)

for n, contour in enumerate(contourList):

axes.plot(contour[:, 1], contour[:, 0], linewidth = 2)

axes.axis("image")

axes.set_xticks([])

axes.set_yticks([])

pp.show()

source.png

[PYTHON/SKIMAGE] 이미지 이어 붙이기(Image Stitching)

■ 이미지를 이어 붙이는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
import numpy as np
import skimage.color as color
import skimage.feature as feature
import skimage.io as io
import skimage.measure as measure
import skimage.transform as transform

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray1 = io.imread("source1.png")
imageNDArray2 = io.imread("source2.png")

orb = feature.ORB(n_keypoints = 1000, fast_threshold = 0.05)

orb.detect_and_extract(imageNDArray1)

keypointNDArray1   = orb.keypoints
descriptorNDArray1 = orb.descriptors

orb.detect_and_extract(imageNDArray2)

keypointNDArray2   = orb.keypoints
descriptorNDArray2 = orb.descriptors

matchNDArray = feature.match_descriptors(descriptorNDArray1, descriptorNDArray2, cross_check = True)

sourceNDArray = keypointNDArray2[matchNDArray[:, 1]][:, ::-1]
targetNDArray = keypointNDArray1[matchNDArray[:, 0]][:, ::-1]

projectiveTransform, inlierNDArray = measure.ransac((sourceNDArray, targetNDArray), transform.ProjectiveTransform, min_samples = 4, residual_threshold = 2)

y, x = imageNDArray2.shape[:2]

cornerNDArray       = np.array([[0, 0], [0, y], [x, 0],[x, y]])
warpedCornerNDArray = projectiveTransform(cornerNDArray)
allCornerNDArray    = np.vstack((warpedCornerNDArray, cornerNDArray))

minimumCornerNDArray = np.min(allCornerNDArray, axis = 0)
maximumCornerNDArray = np.max(allCornerNDArray, axis = 0)

outputShapeNDArray = (maximumCornerNDArray - minimumCornerNDArray)
outputShapeNDArray = np.ceil(outputShapeNDArray[::-1])

similarityTransform = transform.SimilarityTransform(translation = -minimumCornerNDArray)

warpImageNDArray1 = transform.warp(imageNDArray1, similarityTransform.inverse, output_shape = outputShapeNDArray, cval = -1)
warpImageNDArray2 = transform.warp(imageNDArray2, (projectiveTransform + similarityTransform).inverse, output_shape = outputShapeNDArray, cval = -1)

maskImageNDArray1 = (warpImageNDArray1 != -1)

warpImageNDArray1[~maskImageNDArray1] = 0

alphaImageNDArray1 = np.dstack((color.gray2rgb(warpImageNDArray1), maskImageNDArray1))

maskImageNDArray2 = (warpImageNDArray2 != -1)

warpImageNDArray2[~maskImageNDArray2] = 0

alphaImageNDArray2 = np.dstack((color.gray2rgb(warpImageNDArray2), maskImageNDArray2))

mergeImageNDArray = (alphaImageNDArray1 + alphaImageNDArray2)

alphaImageNDArray = mergeImageNDArray[..., 3]

mergeImageNDArray /= np.maximum(alphaImageNDArray, 1)[..., np.newaxis]

io.imshow(mergeImageNDArray)

io.show()

import matplotlib

import numpy as np

import skimage.color as color

import skimage.feature as feature

import skimage.io as io

import skimage.measure as measure

import skimage.transform as transform

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray1 = io.imread("source1.png")

imageNDArray2 = io.imread("source2.png")

orb = feature.ORB(n_keypoints = 1000, fast_threshold = 0.05)

orb.detect_and_extract(imageNDArray1)

keypointNDArray1 = orb.keypoints

descriptorNDArray1 = orb.descriptors

orb.detect_and_extract(imageNDArray2)

keypointNDArray2 = orb.keypoints

descriptorNDArray2 = orb.descriptors

matchNDArray = feature.match_descriptors(descriptorNDArray1, descriptorNDArray2, cross_check = True)

sourceNDArray = keypointNDArray2[matchNDArray[:, 1]][:, ::-1]

targetNDArray = keypointNDArray1[matchNDArray[:, 0]][:, ::-1]

projectiveTransform, inlierNDArray = measure.ransac((sourceNDArray, targetNDArray), transform.ProjectiveTransform, min_samples = 4, residual_threshold = 2)

y, x = imageNDArray2.shape[:2]

cornerNDArray = np.array([[0, 0], [0, y], [x, 0],[x, y]])

warpedCornerNDArray = projectiveTransform(cornerNDArray)

allCornerNDArray = np.vstack((warpedCornerNDArray, cornerNDArray))

minimumCornerNDArray = np.min(allCornerNDArray, axis = 0)

maximumCornerNDArray = np.max(allCornerNDArray, axis = 0)

outputShapeNDArray = (maximumCornerNDArray - minimumCornerNDArray)

outputShapeNDArray = np.ceil(outputShapeNDArray[::-1])

similarityTransform = transform.SimilarityTransform(translation = -minimumCornerNDArray)

warpImageNDArray1 = transform.warp(imageNDArray1, similarityTransform.inverse, output_shape = outputShapeNDArray, cval = -1)

warpImageNDArray2 = transform.warp(imageNDArray2, (projectiveTransform + similarityTransform).inverse, output_shape = outputShapeNDArray, cval = -1)

maskImageNDArray1 = (warpImageNDArray1 != -1)

warpImageNDArray1[~maskImageNDArray1] = 0

alphaImageNDArray1 = np.dstack((color.gray2rgb(warpImageNDArray1), maskImageNDArray1))

maskImageNDArray2 = (warpImageNDArray2 != -1)

warpImageNDArray2[~maskImageNDArray2] = 0

alphaImageNDArray2 = np.dstack((color.gray2rgb(warpImageNDArray2), maskImageNDArray2))

mergeImageNDArray = (alphaImageNDArray1 + alphaImageNDArray2)

alphaImageNDArray = mergeImageNDArray[..., 3]

mergeImageNDArray /= np.maximum(alphaImageNDArray, 1)[..., np.newaxis]

io.imshow(mergeImageNDArray)

io.show()

source1.png source2.png

[PYTHON/SKIMAGE] ORB(Oriented FAST and Rotated BRIEF) 분류기 사용하기

■ ORB(Oriented FAST and Rotated BRIEF) 분류기를 사용하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
import matplotlib.pyplot as pp
import skimage.color as color
import skimage.data as data
import skimage.feature as feature
import skimage.io as io
import skimage.transform as transform

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray          = data.astronaut()
grayscaleImageNDArray = color.rgb2gray(imageNDArray)
rotateImageNDArray    = transform.rotate(grayscaleImageNDArray, 180)
affineTransform       = transform.AffineTransform(scale=(1.3, 1.1), rotation = 0.5, translation = (0, -200))
transformImageNDArray = transform.warp(grayscaleImageNDArray, affineTransform)

orb = feature.ORB(n_keypoints = 200)

orb.detect_and_extract(grayscaleImageNDArray)

grayscaleKeypointNDArray   = orb.keypoints
grayscaleDescriptorNDArray = orb.descriptors

orb.detect_and_extract(rotateImageNDArray)

rotateKeypointNDArray   = orb.keypoints
rotateDescriptorNDArray = orb.descriptors

orb.detect_and_extract(transformImageNDArray)

transformKeyPointNDArray   = orb.keypoints
transformDescriptorNDArray = orb.descriptors

grayscaleRotateMatchNDArray    = feature.match_descriptors(grayscaleDescriptorNDArray, rotateDescriptorNDArray   , cross_check = True)
grayscaleTransformMatchNDArray = feature.match_descriptors(grayscaleDescriptorNDArray, transformDescriptorNDArray, cross_check = True)

figure, axesNDArray = pp.subplots(nrows = 2, ncols = 1)

pp.gray()

feature.plot_matches(axesNDArray[0], grayscaleImageNDArray, rotateImageNDArray, grayscaleKeypointNDArray, rotateKeypointNDArray, grayscaleRotateMatchNDArray)

axesNDArray[0].axis("off")
axesNDArray[0].set_title("Original Image vs. Transformed Image")

feature.plot_matches(axesNDArray[1], grayscaleImageNDArray, transformImageNDArray, grayscaleKeypointNDArray, transformKeyPointNDArray, grayscaleTransformMatchNDArray)

axesNDArray[1].axis("off")
axesNDArray[1].set_title("Original Image vs. Transformed Image")

pp.show()

import matplotlib

import matplotlib.pyplot as pp

import skimage.color as color

import skimage.data as data

import skimage.feature as feature

import skimage.io as io

import skimage.transform as transform

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = data.astronaut()

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

rotateImageNDArray = transform.rotate(grayscaleImageNDArray, 180)

affineTransform = transform.AffineTransform(scale=(1.3, 1.1), rotation = 0.5, translation = (0, -200))

transformImageNDArray = transform.warp(grayscaleImageNDArray, affineTransform)

orb = feature.ORB(n_keypoints = 200)

orb.detect_and_extract(grayscaleImageNDArray)

grayscaleKeypointNDArray = orb.keypoints

grayscaleDescriptorNDArray = orb.descriptors

orb.detect_and_extract(rotateImageNDArray)

rotateKeypointNDArray = orb.keypoints

rotateDescriptorNDArray = orb.descriptors

orb.detect_and_extract(transformImageNDArray)

transformKeyPointNDArray = orb.keypoints

transformDescriptorNDArray = orb.descriptors

grayscaleRotateMatchNDArray = feature.match_descriptors(grayscaleDescriptorNDArray, rotateDescriptorNDArray , cross_check = True)

grayscaleTransformMatchNDArray = feature.match_descriptors(grayscaleDescriptorNDArray, transformDescriptorNDArray, cross_check = True)

figure, axesNDArray = pp.subplots(nrows = 2, ncols = 1)

pp.gray()

feature.plot_matches(axesNDArray[0], grayscaleImageNDArray, rotateImageNDArray, grayscaleKeypointNDArray, rotateKeypointNDArray, grayscaleRotateMatchNDArray)

axesNDArray[0].axis("off")

axesNDArray[0].set_title("Original Image vs. Transformed Image")

feature.plot_matches(axesNDArray[1], grayscaleImageNDArray, transformImageNDArray, grayscaleKeypointNDArray, transformKeyPointNDArray, grayscaleTransformMatchNDArray)

axesNDArray[1].axis("off")

axesNDArray[1].set_title("Original Image vs. Transformed Image")

pp.show()

[PYTHON/SKIMAGE] 해리스 코너 검출기 사용하기

■ 해리스 코너 검출기를 사용하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
import matplotlib.pyplot as pp
import skimage.color as color
import skimage.data as data
import skimage.feature as feature
import skimage.io as io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

cornerNDArray = feature.corner_harris(grayscaleImageNDArray)

coordNDArray = feature.corner_peaks(cornerNDArray, min_distance = 5)

coordSubpixNDArray = feature.corner_subpix(grayscaleImageNDArray, coordNDArray, window_size = 13)

figure, axesSubplot = pp.subplots()

axesSubplot.imshow(grayscaleImageNDArray, interpolation = "nearest", cmap = pp.cm.gray)

axesSubplot.plot(coordNDArray[:, 1], coordNDArray[:, 0], ".b", markersize = 3)
axesSubplot.plot(coordSubpixNDArray[:, 1], coordSubpixNDArray[:, 0], "+r", markersize = 15)

pp.show()

import matplotlib

import matplotlib.pyplot as pp

import skimage.color as color

import skimage.data as data

import skimage.feature as feature

import skimage.io as io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

cornerNDArray = feature.corner_harris(grayscaleImageNDArray)

coordNDArray = feature.corner_peaks(cornerNDArray, min_distance = 5)

coordSubpixNDArray = feature.corner_subpix(grayscaleImageNDArray, coordNDArray, window_size = 13)

figure, axesSubplot = pp.subplots()

axesSubplot.imshow(grayscaleImageNDArray, interpolation = "nearest", cmap = pp.cm.gray)

axesSubplot.plot(coordNDArray[:, 1], coordNDArray[:, 0], ".b", markersize = 3)

axesSubplot.plot(coordSubpixNDArray[:, 1], coordSubpixNDArray[:, 0], "+r", markersize = 15)

pp.show()

[PYTHON/SKIMAGE] probabilistic_hough_line 함수 : 확률적 Hough 변환을 사용해 직선 정보 구하기

■ probabilistic_hough_line 함수에서 확률적 Hough 변환을 사용해 직선 정보를 구하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import color
from skimage import feature
from skimage import io
from skimage import transform

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

edgeImageNDArray = feature.canny(grayscaleImageNDArray, 2, 1, 25)

lineList = transform.probabilistic_hough_line(edgeImageNDArray, threshold = 10, line_length = 5, line_gap = 3)

print(type(lineList))

"""
<class "list">
"""

import matplotlib

from skimage import color

from skimage import feature

from skimage import io

from skimage import transform

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

edgeImageNDArray = feature.canny(grayscaleImageNDArray, 2, 1, 25)

lineList = transform.probabilistic_hough_line(edgeImageNDArray, threshold = 10, line_length = 5, line_gap = 3)

print(type(lineList))

"""

"""

[PYTHON/SKIMAGE] canny 함수 : 캐니 에지 검출기 사용하기

■ canny 함수에서 캐니 에지 검출기를 사용하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import color
from skimage import feature
from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

canyImageNDArray = feature.canny(grayscaleImageNDArray, 3)

io.imshow(canyImageNDArray)

io.show()

import matplotlib

from skimage import color

from skimage import feature

from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

canyImageNDArray = feature.canny(grayscaleImageNDArray, 3)

io.imshow(canyImageNDArray)

io.show()

[PYTHON/SKIMAGE] hough_line 함수 : Hough 변환을 사용해 직선 정보 구하기

■ hough_line 함수에서 Hough 변환을 사용해 직선 정보를 구하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import color
from skimage import feature
from skimage import io
from skimage import transform

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

lineTuple = transform.hough_line(grayscaleImageNDArray)

print(lineTuple)

"""
(array([[0, 0, 0, ..., 0, 0, 0],
       [0, 0, 0, ..., 0, 0, 0],
       [0, 0, 0, ..., 0, 0, 0],
       ...,
       [0, 0, 0, ..., 0, 0, 0],
       [0, 0, 0, ..., 0, 0, 0],
       [0, 0, 0, ..., 0, 0, 0]], dtype=uint64), array([-1.57079633, -1.55334303, -1.53588974, -1.51843645, -1.50098316,
       -1.48352986, -1.46607657, -1.44862328, -1.43116999, -1.41371669,
       -1.3962634 , -1.37881011, -1.36135682, -1.34390352, -1.32645023,
       -1.30899694, -1.29154365, -1.27409035, -1.25663706, -1.23918377,
       -1.22173048, -1.20427718, -1.18682389, -1.1693706 , -1.15191731,
       -1.13446401, -1.11701072, -1.09955743, -1.08210414, -1.06465084,
       -1.04719755, -1.02974426, -1.01229097, -0.99483767, -0.97738438,
       -0.95993109, -0.9424778 , -0.9250245 , -0.90757121, -0.89011792,
       -0.87266463, -0.85521133, -0.83775804, -0.82030475, -0.80285146,
       -0.78539816, -0.76794487, -0.75049158, -0.73303829, -0.71558499,
       -0.6981317 , -0.68067841, -0.66322512, -0.64577182, -0.62831853,
       -0.61086524, -0.59341195, -0.57595865, -0.55850536, -0.54105207,
       -0.52359878, -0.50614548, -0.48869219, -0.4712389 , -0.45378561,
       -0.43633231, -0.41887902, -0.40142573, -0.38397244, -0.36651914,
       -0.34906585, -0.33161256, -0.31415927, -0.29670597, -0.27925268,
       -0.26179939, -0.2443461 , -0.2268928 , -0.20943951, -0.19198622,
       -0.17453293, -0.15707963, -0.13962634, -0.12217305, -0.10471976,
       -0.08726646, -0.06981317, -0.05235988, -0.03490659, -0.01745329,
        0.        ,  0.01745329,  0.03490659,  0.05235988,  0.06981317,
        0.08726646,  0.10471976,  0.12217305,  0.13962634,  0.15707963,
        0.17453293,  0.19198622,  0.20943951,  0.2268928 ,  0.2443461 ,
        0.26179939,  0.27925268,  0.29670597,  0.31415927,  0.33161256,
        0.34906585,  0.36651914,  0.38397244,  0.40142573,  0.41887902,
        0.43633231,  0.45378561,  0.4712389 ,  0.48869219,  0.50614548,
        0.52359878,  0.54105207,  0.55850536,  0.57595865,  0.59341195,
        0.61086524,  0.62831853,  0.64577182,  0.66322512,  0.68067841,
        0.6981317 ,  0.71558499,  0.73303829,  0.75049158,  0.76794487,
        0.78539816,  0.80285146,  0.82030475,  0.83775804,  0.85521133,
        0.87266463,  0.89011792,  0.90757121,  0.9250245 ,  0.9424778 ,
        0.95993109,  0.97738438,  0.99483767,  1.01229097,  1.02974426,
        1.04719755,  1.06465084,  1.08210414,  1.09955743,  1.11701072,
        1.13446401,  1.15191731,  1.1693706 ,  1.18682389,  1.20427718,
        1.22173048,  1.23918377,  1.25663706,  1.27409035,  1.29154365,
        1.30899694,  1.32645023,  1.34390352,  1.36135682,  1.37881011,
        1.3962634 ,  1.41371669,  1.43116999,  1.44862328,  1.46607657,
        1.48352986,  1.50098316,  1.51843645,  1.53588974,  1.55334303]), array([-1082., -1081., -1080., ...,  1080.,  1081.,  1082.]))
"""

import matplotlib

from skimage import color

from skimage import feature

from skimage import io

from skimage import transform

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

lineTuple = transform.hough_line(grayscaleImageNDArray)

print(lineTuple)

"""

(array([[0, 0, 0, ..., 0, 0, 0],

[0, 0, 0, ..., 0, 0, 0],

...,

[0, 0, 0, ..., 0, 0, 0],

[0, 0, 0, ..., 0, 0, 0]], dtype=uint64), array([-1.57079633, -1.55334303, -1.53588974, -1.51843645, -1.50098316,

-1.48352986, -1.46607657, -1.44862328, -1.43116999, -1.41371669,

-1.3962634 , -1.37881011, -1.36135682, -1.34390352, -1.32645023,

-1.30899694, -1.29154365, -1.27409035, -1.25663706, -1.23918377,

-1.22173048, -1.20427718, -1.18682389, -1.1693706 , -1.15191731,

-1.13446401, -1.11701072, -1.09955743, -1.08210414, -1.06465084,

-1.04719755, -1.02974426, -1.01229097, -0.99483767, -0.97738438,

-0.95993109, -0.9424778 , -0.9250245 , -0.90757121, -0.89011792,

-0.87266463, -0.85521133, -0.83775804, -0.82030475, -0.80285146,

-0.78539816, -0.76794487, -0.75049158, -0.73303829, -0.71558499,

-0.6981317 , -0.68067841, -0.66322512, -0.64577182, -0.62831853,

-0.61086524, -0.59341195, -0.57595865, -0.55850536, -0.54105207,

-0.52359878, -0.50614548, -0.48869219, -0.4712389 , -0.45378561,

-0.43633231, -0.41887902, -0.40142573, -0.38397244, -0.36651914,

-0.34906585, -0.33161256, -0.31415927, -0.29670597, -0.27925268,

-0.26179939, -0.2443461 , -0.2268928 , -0.20943951, -0.19198622,

-0.17453293, -0.15707963, -0.13962634, -0.12217305, -0.10471976,

-0.08726646, -0.06981317, -0.05235988, -0.03490659, -0.01745329,

0. , 0.01745329, 0.03490659, 0.05235988, 0.06981317,

0.08726646, 0.10471976, 0.12217305, 0.13962634, 0.15707963,

0.17453293, 0.19198622, 0.20943951, 0.2268928 , 0.2443461 ,

0.26179939, 0.27925268, 0.29670597, 0.31415927, 0.33161256,

0.34906585, 0.36651914, 0.38397244, 0.40142573, 0.41887902,

0.43633231, 0.45378561, 0.4712389 , 0.48869219, 0.50614548,

0.52359878, 0.54105207, 0.55850536, 0.57595865, 0.59341195,

0.61086524, 0.62831853, 0.64577182, 0.66322512, 0.68067841,

0.6981317 , 0.71558499, 0.73303829, 0.75049158, 0.76794487,

0.78539816, 0.80285146, 0.82030475, 0.83775804, 0.85521133,

0.87266463, 0.89011792, 0.90757121, 0.9250245 , 0.9424778 ,

0.95993109, 0.97738438, 0.99483767, 1.01229097, 1.02974426,

1.04719755, 1.06465084, 1.08210414, 1.09955743, 1.11701072,

1.13446401, 1.15191731, 1.1693706 , 1.18682389, 1.20427718,

1.22173048, 1.23918377, 1.25663706, 1.27409035, 1.29154365,

1.30899694, 1.32645023, 1.34390352, 1.36135682, 1.37881011,

1.3962634 , 1.41371669, 1.43116999, 1.44862328, 1.46607657,

1.48352986, 1.50098316, 1.51843645, 1.53588974, 1.55334303]), array([-1082., -1081., -1080., ..., 1080., 1081., 1082.]))

"""

[PYTHON/SKIMAGE] sobel 함수 : 소벨 에지 검출기 사용하기

■ sobel 함수에서 소벨 에지 검출기를 사용하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import color
from skimage import filters
from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

sobelImageNDArray = filters.sobel(grayscaleImageNDArray)

io.imshow(sobelImageNDArray)

io.show()

import matplotlib

from skimage import color

from skimage import filters

from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

sobelImageNDArray = filters.sobel(grayscaleImageNDArray)

io.imshow(sobelImageNDArray)

io.show()

[PYTHON/SKIMAGE] threshold_otsu 함수 : 이미지 임계 처리하기

■ threshold_otsu 함수를 사용해 이미지 임계를 처리하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import color
from skimage import filters
from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

threshValue = filters.threshold_otsu(grayscaleImageNDArray)

threshImageNDArray = grayscaleImageNDArray > threshValue

io.imshow(threshImageNDArray)

io.show()

import matplotlib

from skimage import color

from skimage import filters

from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

threshValue = filters.threshold_otsu(grayscaleImageNDArray)

threshImageNDArray = grayscaleImageNDArray > threshValue

io.imshow(threshImageNDArray)

io.show()

[PYTHON/SKIMAGE] binary_dilation 함수 : 팽창(Dilation) 필터 사용하기

■ binary_dilation 함수에서 팽창(Dilation) 필터를 사용하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import color
from skimage import data
from skimage import io
from skimage import morphology

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

dilationImageNDArray = morphology.binary_dilation(grayscaleImageNDArray)

io.imshow(dilationImageNDArray)

io.show()

import matplotlib

from skimage import color

from skimage import data

from skimage import io

from skimage import morphology

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

dilationImageNDArray = morphology.binary_dilation(grayscaleImageNDArray)

io.imshow(dilationImageNDArray)

io.show()

[PYTHON/SKIMAGE] binary_erosion 함수 : 침식(Erosion) 필터 사용하기

■ binary_erosion 함수에서 침식(Erosion) 필터를 사용하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import color
from skimage import data
from skimage import io
from skimage import morphology

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

erosionImageNDArray = morphology.binary_erosion(grayscaleImageNDArray)

io.imshow(erosionImageNDArray)

io.show()

import matplotlib

from skimage import color

from skimage import data

from skimage import io

from skimage import morphology

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

erosionImageNDArray = morphology.binary_erosion(grayscaleImageNDArray)

io.imshow(erosionImageNDArray)

io.show()

[PYTHON/SKIMAGE] imread 함수 : 이미지 로드하기

■ imread 함수를 사용해 이미지를 로드하는 방법을 보여준다. ▶ main.py


import matplotlib
from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("sample.png")

io.imshow(imageNDArray)

io.show()

import matplotlib

from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("sample.png")

io.imshow(imageNDArray)

io.show()

▶ requirements.txt


cycler==0.11.0
fonttools==4.34.4
imageio==2.21.1
kiwisolver==1.4.4
matplotlib==3.5.3
networkx==2.8.5
numpy==1.23.1
packaging==21.3
Pillow==9.2.0
pip==22.0.4
pyparsing==3.0.9
python-dateutil==2.8.2
PyWavelets==1.3.0
scikit-image==0.19.3
scipy==1.9.0
setuptools==58.1.0
six==1.16.0
tifffile==2022.8.8

cycler==0.11.0

fonttools==4.34.4

imageio==2.21.1

kiwisolver==1.4.4

matplotlib==3.5.3

networkx==2.8.5

numpy==1.23.1

packaging==21.3

Pillow==9.2.0

pip==22.0.4

pyparsing==3.0.9

python-dateutil==2.8.2

PyWavelets==1.3.0

scikit-image==0.19.3

scipy==1.9.0

setuptools==58.1.0

six==1.16.0

tifffile==2022.8.8

sample.png

[PYTHON/SKIMAGE] median 함수 : 중간값(Median) 필터 사용하기

■ median 함수에서 중간값(Median) 필터를 사용하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import color
from skimage import filters
from skimage import io
from skimage.morphology import disk

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("sample.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

medianNDArray = filters.median(grayscaleImageNDArray, disk(7))

io.imshow(medianNDArray)

io.show()

import matplotlib

from skimage import color

from skimage import filters

from skimage import io

from skimage.morphology import disk

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("sample.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

medianNDArray = filters.median(grayscaleImageNDArray, disk(7))

io.imshow(medianNDArray)

io.show()

※ imageNDArray : 3차원 배열 ※ grayscaleImageNDArray : 2차원

[PYTHON/SKIMAGE] gaussian 함수 : 가우시안 블러(Gaussian Blur) 필터 사용하기

■ gaussian 함수에서 가우시안 블러(Gaussian Blur) 필터를 사용하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import filters
from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

gaussianImageNDArray = filters.gaussian(imageNDArray, sigma = 5)

io.imshow(gaussianImageNDArray)

io.show()

import matplotlib

from skimage import filters

from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

gaussianImageNDArray = filters.gaussian(imageNDArray, sigma = 5)

io.imshow(gaussianImageNDArray)

io.show()

[PYTHON/SKIMAGE] polygon 함수 : 다각형 그리기

■ polygon 함수를 사용해 다각형을 그리는 방법을 보여준다. ▶ 예제 코드 (PY)


import numpy as np
import matplotlib
from skimage import io, draw

matplotlib.use('TkAgg')

io.use_plugin('matplotlib')

imageNDArray = np.zeros((100, 100), dtype = np.uint8)

xNDArray = np.array([10, 60, 40, 10])
yNDArray = np.array([10, 25, 80, 50])

x, y = draw.polygon(yNDArray, xNDArray)

imageNDArray[x, y] = 1

io.imshow(imageNDArray)

io.show()

import numpy as np

import matplotlib

from skimage import io, draw

matplotlib.use('TkAgg')

io.use_plugin('matplotlib')

imageNDArray = np.zeros((100, 100), dtype = np.uint8)

xNDArray = np.array([10, 60, 40, 10])

yNDArray = np.array([10, 25, 80, 50])

x, y = draw.polygon(yNDArray, xNDArray)

imageNDArray[x, y] = 1

io.imshow(imageNDArray)

io.show()

[PYTHON/SKIMAGE] ellipse 함수 : 타원 그리기

■ ellipse 함수를 사용해 타원을 그리는 방법을 보여준다. ▶ 예제 코드 (PY)


import numpy as np
import matplotlib
from skimage import io, draw

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = np.zeros((100, 100), dtype = np.uint8)

x, y = draw.ellipse(50, 50, 10, 20)

imageNDArray[x, y] = 1

io.imshow(imageNDArray)

io.show()

import numpy as np

import matplotlib

from skimage import io, draw

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = np.zeros((100, 100), dtype = np.uint8)

x, y = draw.ellipse(50, 50, 10, 20)

imageNDArray[x, y] = 1

io.imshow(imageNDArray)

io.show()

[PYTHON/SKIMAGE] circle_perimeter 함수 : 원 그리기

■ circle_perimeter 함수를 사용해 원을 그리는 방법을 보여준다. ▶ 예제 코드 (PY)


import numpy as np
import matplotlib
from skimage import io, draw

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = np.zeros((100, 100), dtype = np.uint8)

x, y = draw.circle_perimeter(50, 50, 10)

imageNDArray[x, y] = 1

io.imshow(imageNDArray)

io.show()

import numpy as np

import matplotlib

from skimage import io, draw

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = np.zeros((100, 100), dtype = np.uint8)

x, y = draw.circle_perimeter(50, 50, 10)

imageNDArray[x, y] = 1

io.imshow(imageNDArray)

io.show()

[PYTHON/SKIMAGE] rgb2hsv 함수 : HSV 색상 모드 이미지 구하기

■ rgb2hsv 함수를 사용해 HSV 색상 모드 이미지를 구하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import io, color

matplotlib.use('TkAgg')

io.use_plugin('matplotlib')

imageNDArray = io.imread("source.png")

hsvImageNDArray = color.rgb2hsv(imageNDArray)

io.imshow(hsvImageNDArray)

io.show()

import matplotlib

from skimage import io, color

matplotlib.use('TkAgg')

io.use_plugin('matplotlib')

imageNDArray = io.imread("source.png")

hsvImageNDArray = color.rgb2hsv(imageNDArray)

io.imshow(hsvImageNDArray)

io.show()

[PYTHON/SKIMAGE] rgb2gray 함수 : 회색조 색상 모드 이미지 구하기

■ rgb2gray 함수를 사용해 회색조 색상 모드 이미지를 구하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import io, color

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

io.imshow(grayscaleImageNDArray)

io.show()

import matplotlib

from skimage import io, color

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

grayscaleImageNDArray = color.rgb2gray(imageNDArray)

io.imshow(grayscaleImageNDArray)

io.show()

[PYTHON/SKIMAGE] imsave 함수 : 이미지 저장하기

■ imsave 함수를 사용해 이미지를 저장하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

io.imsave("target.png", imageNDArray)

import matplotlib

from skimage import io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

imageNDArray = io.imread("source.png")

io.imsave("target.png", imageNDArray)

[PYTHON/SKIMAGE] 내장된 이미지 사용하기

■ 내장된 이미지를 사용하는 방법을 보여준다. ▶ 예제 코드 (PY)


import matplotlib
from skimage import data, io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

io.imshow(data.camera())

io.show()

import matplotlib

from skimage import data, io

matplotlib.use("TkAgg")

io.use_plugin("matplotlib")

io.imshow(data.camera())

io.show()