形态学方法
当图像经过预处理进行增强和阈值等性能操作时,图像就有可能得到一些噪声。从而导致图像中存在像素信息不平衡的问题。
形态学的操作主要是去除影响图像形状和信息的噪声。形态学运算在图像分割中非常有用,可以得到无噪声的二值图像。
基本的形态操作是侵蚀和膨胀。下面对这两种操作进行说明:
膨胀
在放大操作中,如果物体是白色的,那么白色像素周围的像素就会增大。它增加的区域取决于物体像素的形状。膨胀过程增加了对象的像素数,减少了非对象的像素数。
具有不同内核大小和迭代的膨胀的Python代码
import numpy as np
import imutils
import cv2#reading the input image
img = cv2.imread('thumb.png') #reads the image
#cv2.imwrite('Input_image.jpg',image)
#Resizing the image
scale_percent = 70
width = int(img.shape[1] * scale_percent / 100)
height = int(img.shape[0] * scale_percent / 100)
dim = (width, height)
# resize the input image
image = cv2.resize(img, dim, interpolation = cv2.INTER_AREA)
kernel = np.ones((1,1), dtype = "uint8")/9
dilation = cv2.dilate(image,kernel,iterations = 1)
cv2.imwrite('dilation.jpg', dilation)
kernel = np.ones((2,2), dtype = "uint8")/9
dilation = cv2.dilate(image,kernel,iterations = 1)
cv2.imwrite('dilation.jpg', dilation)
kernel = np.ones((2,2), dtype = "uint8")/9
dilation = cv2.dilate(image,kernel,iterations = 3)
cv2.imwrite('dilation.jpg', dilation)
kernel = np.ones((2,2), dtype = "uint8")/9
dilation = cv2.dilate(image,kernel,iterations = 5)
cv2.imwrite('dilation.jpg', dilation)
kernel = np.ones((3,3), dtype = "uint8")/9
dilation = cv2.dilate(image,kernel,iterations = 2)
cv2.imwrite('dilation.jpg', dilation)
侵蚀
侵蚀函数正好与膨胀功函数相反。侵蚀作用使物体体积变小。侵蚀过程增加了非目标像素,减少了目标像素。
具有不同内核大小和迭代的侵蚀的Python代码
import numpy as np
import imutils
import cv2
#reading the input image
img = cv2.imread('thumb.png')
#cv2.imwrite('Input_image.jpg',image)
#Resizing the image
scale_percent = 70
width = int(img.shape[1] * scale_percent / 100)
height = int(img.shape[0] * scale_percent / 100)
dim = (width, height)
# resize the input image
image = cv2.resize(img, dim, interpolation = cv2.INTER_AREA)
kernel = np.ones((1,1), dtype = "uint8")/9
erosion = cv2.erode(image, kernel, iterations = 1)
cv2.imwrite('erosion.jpg', erosion)
kernel = np.ones((2,2), dtype = "uint8")/9
erosion = cv2.erode(image, kernel, iterations = 2)
cv2.imwrite('erosion.jpg', erosion)
kernel = np.ones((2,2), dtype = "uint8")/9
erosion = cv2.erode(image, kernel, iterations = 3)
cv2.imwrite('erosion.jpg', erosion)
kernel = np.ones((2,2), dtype = "uint8")/9
erosion = cv2.erode(image, kernel, iterations = 5)
cv2.imwrite('erosion.jpg', erosion)
kernel = np.ones((5,5), dtype = "uint8")/9
erosion = cv2.erode(image, kernel, iterations = 2)
cv2.imwrite('erosion.jpg', erosion)
开操作
此方法可用于从图像中去除噪声。该方法的工作功能是先腐蚀再膨胀,以保持物体像素的原始性,去除背景中的小噪声。
import numpy as np
import imutils
import cv2
#reading the input image
img = cv2.imread('11.png')
kernel = np.ones((5,5), dtype = "uint8")/9
opening = cv2.morphologyEx(img, cv2.MORPH_OPEN, kernel)
cv2.imwrite('opening.jpg', opening)
闭操作
此方法可用于从图像中去除噪声。这种方法的工作功能是先膨胀再腐蚀,去除内部的小噪声。
import numpy as np
import imutils
import cv2
#reading the input image
img = cv2.imread('thumb.png')
kernel = np.ones((9,9), dtype = "uint8")/9
closing = cv2.morphologyEx(img, cv2.MORPH_CLOSE, kernel)
cv2.imwrite('closing.jpg', closing)
形态学梯度
这种方法是膨胀图与腐蚀图之差。
import numpy as np
import imutils
import cv2
#reading the input image
img = cv2.imread('g1.png')
kernel = np.ones((6,6), dtype = "uint8")/9
gradient = cv2.morphologyEx(img, cv2.MORPH_GRADIENT, kernel)
cv2.imwrite('gradient.jpg', gradient)
总结
这些操作是处理二进制图像的一种非常简单的方法,也是图像处理应用程序中预处理的一部分。
作者:Amit Chauhan