@@ -3,5 +3,6 @@
 *.png
 *.zip
 *.txt
+*.rar
 *.dat
 /charts
@@ -2,24 +2,49 @@
 import mahotas as mh
 image = mh.imread('scene00.jpg')
 from matplotlib import pyplot as plt
-plt.imshow(image)
-plt.show()
+import pylab
+import sys
+
+
+# plt.imshow(image)
+# plt.show()
+
+
+#matplotlib默认将单通道图像转换成了假彩色图像，较高值用红色，较低值用蓝色
 image = mh.colors.rgb2grey(image, dtype=np.uint8)
 plt.imshow(image) # Display the image
+
+#现在图像变成灰度图了
 plt.gray()
+#将图像输入给otsu方法，该方法会找到合适的域值
 thresh = mh.thresholding.otsu(image)
 print('Otsu threshold is {}.'.format(thresh))
 # Otsu threshold is 138.
+
 plt.imshow(image > thresh)
 
-im16 = mh.gaussian_filter(image,16)
+
+#高斯滤波器，其中16为滤波器的大小（滤波器的标准差），越大越模糊
+im16 = mh.gaussian_filter(image,1)
+
+#换一张新的图片
 im = mh.demos.load('lenna')
 
+#加入椒盐噪声
+salt = np.random.random(im.shape) > .975
+pepper = np.random.random(im.shape) > .975
+im_salt = mh.stretch(im)
+im_salt = np.maximum(salt*170, im_salt)
+im_salt = np.minimum(pepper*30 + im_salt*(~pepper), im_salt)
+plt.imshow(im_salt)
+# plt.show()
+#聚焦中心
 r,g,b = im.transpose(2,0,1) 
 r12 = mh.gaussian_filter(r, 12.)
 g12 = mh.gaussian_filter(g, 12.)
 b12 = mh.gaussian_filter(b, 12.)
 im12 = mh.as_rgb(r12,g12,b12)
+
 h, w = r.shape # height and width
 Y, X = np.mgrid[:h,:w]
 Y = Y-h/2. # center at h/2
@@ -36,11 +61,18 @@
 C = C[:,:,None] # This adds a dummy third dimension to C
 
 ringed = mh.stretch(im*C + (1-C)*im12)
+plt.imshow(ringed)
+# plt.show()
 
+#计算图像特征,haralick_features是一个4*13数组，第一维代表4个可能的方向（上下左右）
 haralick_features = mh.features.haralick(image)
+#如果我们对方向不感兴趣，那么可以对方向（即第一维进行算术平均）
 haralick_features_mean = np.mean(haralick_features, axis=0)
+#将原始向量(4,13)进行拼接，变成一个(52,)的向量
 haralick_features_all = np.ravel(haralick_features)
+print(haralick_features_all.shape) #(52,)
 
+#有了特征我们就可以去对图像进行分类
 from glob import glob
 images = glob('../SimpleImageDataset/*.jpg')
 features = []
@@ -51,23 +83,23 @@
   im = mh.colors.rgb2gray(im, dtype=np.uint8)
   features.append(mh.features.haralick(im).ravel())
 
-features = np.array(features)
-labels = np.array(labels)
-
+features = np.array(features)  #(90,52)
+labels = np.array(labels)      #(90,)
+#训练一个Logistic 回归模型
 from sklearn.pipeline import Pipeline
 from sklearn.preprocessing import StandardScaler
 from sklearn.linear_model import LogisticRegression
 clf = Pipeline([('preproc', StandardScaler()),
                 ('classifier', LogisticRegression())])
 
-
 from sklearn import cross_validation
 cv = cross_validation.LeaveOneOut(len(images))
 scores = cross_validation.cross_val_score(
     clf, features, labels, cv=cv)
 print('Accuracy: {:.1%}'.format(scores.mean()))
 # Accuracy: 81.1%
 
+#增加特征
 def chist(im):
     im = im // 64
     r,g,b = im.transpose((2,0,1))
@@ -82,8 +114,6 @@ def chist(im):
   im = mh.imread(im)
   features.append(chist(im))
 
-
-
 features = []
 for im in images:
   imcolor = mh.imread(im)
@@ -93,13 +123,12 @@ def chist(im):
           chist(imcolor),
       ]))
 
-
 scores = cross_validation.cross_val_score(
     clf, features, labels, cv=cv)
 print('Accuracy: {:.1%}'.format(scores.mean()))
 # Accuracy: 95.6%
 
-
+#忽略边缘的信息
 features = []
 for im in images:
   imcolor = mh.imread(im)
@@ -110,7 +139,7 @@ def chist(im):
           mh.features.haralick(im).ravel(),
           chist(imcolor),
       ]))
-
+#对特征做归一化
 sc = StandardScaler()
 features = sc.fit_transform(features)
 from scipy.spatial import distance
@@ -146,7 +175,6 @@ def chist(im):
 
 
 
-
 from mahotas.features import surf
 image = mh.demos.load('lena')
 image = mh.colors.rgb2gray(image, dtype=np.uint8)
@@ -161,10 +189,13 @@ def chist(im):
   alldescriptors.append(surf.dense(image, spacing=16))
 # get all descriptors into a single array
 concatenated = np.concatenate(alldescriptors)
+print(concatenated.shape)
 print('Number of descriptors: {}'.format(
        len(concatenated)))
 # use only every 64th vector
 concatenated = concatenated[::64] 
+print(concatenated.shape)
+
 from sklearn.cluster import KMeans # FIXME CAPITALIZATION
 k = 256
 km = KMeans(k)
@@ -178,6 +209,7 @@ def chist(im):
   )
 # build single array and convert to float
 features = np.array(features, dtype=float)
+
 scores = cross_validation.cross_val_score(
    clf, features, labels, cv=cv)
 print('Accuracy: {:.1%}'.format(scores.mean()))
 
@@ -2,7 +2,7 @@
 
 mkdir -p AnimTransDistr
 cd AnimTransDistr
-curl -O http://vision.stanford.edu/Datasets/AnimTransDistr.rar
+#curl -O http://vision.stanford.edu/Datasets/AnimTransDistr.rar
 unrar x AnimTransDistr.rar
 # The following file is a weird file:
 rm Anims/104034.jpg
@@ -37,7 +37,7 @@ def images():
     '''
     for ci, cl in enumerate(classes):
         images = glob('{}/{}/*.jpg'.format(basedir, cl))
-        for im in sorted(images):
+        for im in sorted(images[:50]):
             yield im, ci
 
 classes = [
 
@@ -38,8 +38,8 @@
 print('Computing neighbors...')
 sc = StandardScaler()
 features = sc.fit_transform(features)
+#计算距离矩阵
 dists = distance.squareform(distance.pdist(features))
-
 print('Plotting...')
 fig, axes = plt.subplots(2, 9, figsize=(16,8))
 
 
@@ -34,7 +34,7 @@ def plot_correlation_demo():
     np.random.seed(1)  # to reproduce the data later on
     pylab.clf()
     pylab.figure(num=None, figsize=(8, 8))
-    x = np.arange(0, 10, 0.2)
+    x = np.arange(0, 10, 0.2) 
 
     pylab.subplot(221)
     y = 0.5 * x + norm.rvs(1, scale=.01, size=len(x))
 
@@ -30,7 +30,7 @@ def plot_demo_1():
     fig = pylab.figure(figsize=(10, 4))
 
     ax = fig.add_subplot(121, projection='3d')
-    ax.set_facecolor('white')
+    ax.set_facecolor('white') 
     # ax.set_axis_bgcolor('white')
 
     mds = manifold.MDS(n_components=3)
-Original file line number
+Diff line change
 *.png
 *.zip
 *.txt
 +*.rar
 *.dat
 /charts