python的数字图像处理学习（2）

图像的重定义大小，图像的缩扩，图像的旋转：

 from skimage import transform,data
 import matplotlib.pyplot as plt
 img = data.camera()
 print(img.shape)
 plt.subplot(221)
 plt.imshow(img)
 plt.subplot(222)
 plt.imshow(transform.resize(img,(64,64)))
 plt.subplot(223)
 plt.imshow(transform.rescale(img,0.2))
 plt.subplot(224)
 plt.imshow(transform.rotate(img,30,resize=True))
 plt.show()

产生高斯金字塔

 import numpy as np
 import matplotlib.pyplot as plt
 from skimage import data,transform
 image = data.astronaut()  #载入宇航员图片
 pyramid = transform.pyramid_gaussian(image, downscale=2)  #产生高斯金字塔图像
 #pyramid = transform.pyramid_laplacian(image, downscale=2)
 #共生成了log(512)=9幅金字塔图像，加上原始图像共10幅，pyramid[0]-pyramid[1]
 i = 1
 for p in pyramid:
     plt.subplot(2,5,i)
     i+=1
     #p[:,:,:]*=255
     plt.title(p.shape)
     plt.imshow(p)
 plt.show()

gamma调整原理：I=I^g 如果gamma>1, 新图像比原图像暗。如果gamma<1,新图像比原图像亮

log对数调整I=log(I)

对比度是否偏低判断：exposure.is_low_contrast(img)

 from skimage import data, exposure, img_as_float
 import matplotlib.pyplot as plt
 image = img_as_float(data.moon())
 gam1= exposure.adjust_gamma(image, 4)   #调暗
 gam2= exposure.adjust_gamma(image, 0.7)  #调亮
 gam3= exposure.adjust_log(image)   #对数调整
 plt.figure('adjust_gamma',figsize=(10,10))
 plt.subplot(141)
 plt.imshow(image)
 plt.subplot(142)
 plt.imshow(gam1)
 plt.subplot(143)
 plt.imshow(gam2,plt.cm.gray)
 plt.subplot(144)
 plt.imshow(gam3)
 plt.show()   #原理：I=Ig
 result=exposure.is_low_contrast(gam1)
 result

调整图片强度，不是很懂参数...

 import numpy as np
 from skimage import exposure
 image = data.moon()
 mat=exposure.rescale_intensity(image,out_range=(0,100))
 plt.subplot(121)
 plt.imshow(mat)
 print(image)
 print(mat)
 mat1=exposure.rescale_intensity(image, in_range=(0, 200))
 plt.subplot(122)
 plt.imshow(mat1)
 print(mat1.min())
 print(mat1)

绘制直方图

 from skimage import data
 import matplotlib.pyplot as plt
 img=data.camera()
 plt.figure("hist")
 arr=img.flatten()
 n, bins, patches = plt.hist(arr, bins=256, normed=1,facecolor='red')
 plt.show()

彩色图片三通道直方图：

 from skimage import data
 import matplotlib.pyplot as plt
 img=data.astronaut()
 ar=img[:,:,0].flatten()
 plt.hist(ar, bins=256, normed=1,facecolor='r',edgecolor='r',hold=1)
 ag=img[:,:,1].flatten()
 plt.hist(ag, bins=256, normed=1, facecolor='g',edgecolor='g',hold=1)
 ab=img[:,:,2].flatten()
 plt.hist(ab, bins=256, normed=1, facecolor='b',edgecolor='b')
 plt.show()

直方图均衡化exposure.equalize_hist(img)

对图像中像素个数多的灰度级进行展宽，而对图像中像素个数少的灰度进行压缩，从而扩展取值的动态范围，提高了对比度和灰度色调的变化，使图像更加清晰。

 from skimage import data,exposure
 import matplotlib.pyplot as plt
 img=data.moon()
 plt.figure("hist",figsize=(8,8))

 arr=img.flatten()
 plt.subplot(221)
 plt.imshow(img,plt.cm.gray)  #原始图像
 plt.subplot(222)
 plt.hist(arr, bins=256, normed=1,edgecolor='None',facecolor='red') #原始图像直方图

 img1=exposure.equalize_hist(img)
 arr1=img1.flatten()
 plt.subplot(223)
 plt.imshow(img1,plt.cm.gray)  #均衡化图像
 plt.subplot(224)
 arr1*=255
 plt.hist(arr1, bins=256, normed=1,edgecolor='None',facecolor='red') #均衡化直方图

 plt.show()

图像滤波：

平滑滤波，用来抑制噪声；微分算子，可以用来检测边缘和特征提取。

sobel、roberts、scharr、prewitt、canny算子

gabor、gaussian、median滤波

水平、垂直边缘检测

正负交叉边缘检测

 from skimage import data,filters,feature
 import matplotlib.pyplot as plt
 from skimage.morphology import disk
 img = data.camera()
 edges = filters.sobel(img)
 edges = filters.roberts(img)
 edges = filters.scharr(img)
 edges = filters.prewitt(img)
 edges = feature.canny(img,sigma=3)
 edges,filt_imag  = filters.gabor(img, frequency=0.5)
 edges = filters.gaussian(img,sigma=5)
 edges = filters.median(img,disk(9))
 edges = filters.sobel_h(img)
 #水平边缘检测：sobel_h, prewitt_h, scharr_h
 #垂直边缘检测： sobel_v, prewitt_v, scharr_v
 edges = filters.roberts_neg_diag(img)
 edges = filters.roberts_pos_diag(img)
 plt.imshow(edges,plt.cm.gray)

图像阈值判断与分割的各种方法：

 from skimage import data,filters
 import matplotlib.pyplot as plt
 image = data.camera()
 thresh = filters.threshold_otsu(image)
 thresh = filters.threshold_yen(image)
 thresh = filters.threshold_li(image)
 thresh = filters.threshold_isodata(image)

 dst =(image <= thresh)*1.0   #根据阈值进行分割
 #dst =filters.threshold_adaptive(image, 31,'mean')
 plt.subplot(121)
 plt.title('original image')
 plt.imshow(image,plt.cm.gray)
 plt.subplot(122)
 plt.title('binary image')
 plt.imshow(dst,plt.cm.gray)
 plt.show()

图形的绘制，与颜色。有各种各样的图形啊...

 from skimage import draw,data
 import matplotlib.pyplot as plt
 img=data.chelsea()
 rr, cc=draw.ellipse(150, 150, 30, 80)   #返回像素坐标
 draw.set_color(img,[rr,cc],[255,0,0])
 plt.imshow(img,plt.cm.gray)

图像的膨胀，腐蚀

 from skimage import data
 import skimage.morphology as sm
 import matplotlib.pyplot as plt
 img=data.checkerboard()
 dst=sm.dilation(img,sm.square(5))  #用边长为15的正方形滤波器进行膨胀滤波
 dst1=sm.erosion(img,sm.square(5))  #用边长为5的正方形滤波器进行膨胀滤波
 plt.figure(figsize=(8,8))
 plt.subplot(131)
 plt.imshow(img,plt.cm.gray)
 plt.subplot(132)
 plt.imshow(dst,plt.cm.gray)
 plt.subplot(133)
 plt.imshow(dst1,plt.cm.gray)
 #找到像素值为1的点，将它的邻近像素点都设置成这个值。1值表示白，0值表示黑，因此膨胀操作可以扩大白色值范围，压缩黑色值范围。一般用来扩充边缘或填充小的孔洞
 #将0值扩充到邻近像素。扩大黑色部分，减小白色部分。可用来提取骨干信息，去掉毛刺，去掉孤立的像素。

图像开运算，图像闭运算：

 from skimage import io,color,data
 import skimage.morphology as sm
 import matplotlib.pyplot as plt
 img=color.rgb2gray(data.camera())
 dst=sm.opening(img,sm.disk(9))  #用边长为9的圆形滤波器进行膨胀腐蚀滤波
 dst1=sm.closing(img,sm.disk(9))  #用边长为5的圆形滤波器进行腐蚀膨胀滤波
 plt.figure(figsize=(10,10))
 plt.subplot(131)
 plt.imshow(img,plt.cm.gray)
 plt.subplot(132)
 plt.imshow(dst,plt.cm.gray)
 plt.subplot(133)
 plt.imshow(dst1,plt.cm.gray)

白帽（white-tophat)。黑帽（black-tophat)。

 from skimage import io,color
 import skimage.morphology as sm
 import matplotlib.pyplot as plt
 img=color.rgb2gray(data.camera())
 dst=sm.white_tophat(img,sm.square(21))  #将原图像减去它的开运算值，返回比结构化元素小的白点
 dst1=sm.black_tophat(img,sm.square(21))  #将原图像减去它的闭运算值，返回比结构化元素小的黑点，且将这些黑点反色。
 plt.figure('morphology',figsize=(10,10))
 plt.subplot(131)
 plt.imshow(img,plt.cm.gray)
 plt.subplot(132)
 plt.imshow(dst,plt.cm.gray)
 plt.subplot(133)
 plt.imshow(dst1,plt.cm.gray)