如何在MatPlotLib图像中加粗白色区域？

Question

我已经使用MatPlotLib在Python中阅读了一个单词的图片，将其反转，并希望将所有白色区域变大以增强单词的强度。我该怎么办？

我的假设是，我需要查看白色像素 r 范围内的所有像素，并将它们也着色为白色。这是正确的吗？

请在下面找到代码，示例和所需结果。图片的尺寸为（465，748）和

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.image as mpimg
import pandas as pd

# Code to transform RGB-image into GrayScale and subsequently into inverted BW image.
def rgb2gray(rgb):
    # Convert RGB to GrayScale
    gray_value = np.dot(rgb[...,:3], [0.299, 0.587, 0.114])

    # Invert GrayScale.
    gray_value = 1 - gray_value

    # Apply a threshold.
    gray_value[gray_value >= 0.5] = 1
    gray_value[gray_value < 0.5] = 0
    return gray_value

# Read in image.
img = mpimg.imread('Data/Development/Image_Test.png')     

# BW-scale.
gray = rgb2gray(img)    

# Show image.
plt.imshow(gray, cmap = plt.get_cmap('gray'))

源图像：here

期望的结果：

Answer 1

解决方案是让白色区域dilate。在此解决方案中，通过让膨胀运行10次迭代，我得到了 r = 10个像素。

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.image as mpimg
import pandas as pd

# Code to transform RGB-image into GrayScale and subsequently into inverted BW    image.
def rgb2gray(rgb):
    # Convert RGB to GrayScale
    gray_value = np.dot(rgb[...,:3], [0.299, 0.587, 0.114])

    # Invert GrayScale.
    gray_value = 1 - gray_value

    # Apply a threshold.
    gray_value[gray_value >= 0.5] = 1
    gray_value[gray_value < 0.5] = 0
    return gray_value

# Read in image.
img = mpimg.imread('Data/Development/Image_Test.png')     

# BW-scale.
gray = rgb2gray(img)    

# Dilate the white areas.
gray = ndimage.binary_dilation( gray, iterations = 10 ).astype( gray.dtype )

# Show image.
plt.imshow(gray, cmap = plt.get_cmap('gray'))