|
| 1 | +# -*- coding: utf-8 -*- |
| 2 | +import os |
| 3 | +import cv2 |
| 4 | +import numpy as np |
| 5 | +#from sklearn.cross_validation import train_test_split |
| 6 | +from sklearn.model_selection import train_test_split |
| 7 | +from sklearn.metrics import confusion_matrix, classification_report |
| 8 | + |
| 9 | +#---------------------------------------------------------------------------------- |
| 10 | +# 第一步 切分训练集和测试集 |
| 11 | +#---------------------------------------------------------------------------------- |
| 12 | + |
| 13 | +X = [] #定义图像名称 |
| 14 | +Y = [] #定义图像分类类标 |
| 15 | +Z = [] #定义图像像素 |
| 16 | + |
| 17 | +for i in range(0, 10): |
| 18 | + #遍历文件夹，读取图片 |
| 19 | + for f in os.listdir("photo/%s" % i): |
| 20 | + #获取图像名称 |
| 21 | + X.append("photo//" +str(i) + "//" + str(f)) |
| 22 | + #获取图像类标即为文件夹名称 |
| 23 | + Y.append(i) |
| 24 | + |
| 25 | +X = np.array(X) |
| 26 | +Y = np.array(Y) |
| 27 | + |
| 28 | +#随机率为100% 选取其中的30%作为测试集 |
| 29 | +X_train, X_test, y_train, y_test = train_test_split(X, Y, |
| 30 | +test_size=0.3, random_state=1) |
| 31 | + |
| 32 | +print(len(X_train), len(X_test), len(y_train), len(y_test)) |
| 33 | + |
| 34 | +#---------------------------------------------------------------------------------- |
| 35 | +# 第二步 图像读取及转换为像素直方图 |
| 36 | +#---------------------------------------------------------------------------------- |
| 37 | + |
| 38 | +#训练集 |
| 39 | +XX_train = [] |
| 40 | +for i in X_train: |
| 41 | + #读取图像 |
| 42 | + #print i |
| 43 | + image = cv2.imread(i) |
| 44 | + |
| 45 | + #图像像素大小一致 |
| 46 | + img = cv2.resize(image, (256,256), |
| 47 | + interpolation=cv2.INTER_CUBIC) |
| 48 | + |
| 49 | + #计算图像直方图并存储至X数组 |
| 50 | + hist = cv2.calcHist([img], [0,1], None, |
| 51 | + [256,256], [0.0,255.0,0.0,255.0]) |
| 52 | + |
| 53 | + XX_train.append(((hist/255).flatten())) |
| 54 | + |
| 55 | +#测试集 |
| 56 | +XX_test = [] |
| 57 | +for i in X_test: |
| 58 | + #读取图像 |
| 59 | + #print i |
| 60 | + image = cv2.imread(i) |
| 61 | + |
| 62 | + #图像像素大小一致 |
| 63 | + img = cv2.resize(image, (256,256), |
| 64 | + interpolation=cv2.INTER_CUBIC) |
| 65 | + |
| 66 | + #计算图像直方图并存储至X数组 |
| 67 | + hist = cv2.calcHist([img], [0,1], None, |
| 68 | + [256,256], [0.0,255.0,0.0,255.0]) |
| 69 | + |
| 70 | + XX_test.append(((hist/255).flatten())) |
| 71 | + |
| 72 | +#---------------------------------------------------------------------------------- |
| 73 | +# 第三步 基于决策树的图像分类处理 |
| 74 | +#---------------------------------------------------------------------------------- |
| 75 | + |
| 76 | +from sklearn.tree import DecisionTreeClassifier |
| 77 | +clf = DecisionTreeClassifier().fit(XX_train, y_train) |
| 78 | +predictions_labels = clf.predict(XX_test) |
| 79 | + |
| 80 | +print(u'预测结果:') |
| 81 | +print(predictions_labels) |
| 82 | + |
| 83 | +print(u'算法评价:') |
| 84 | +print((classification_report(y_test, predictions_labels))) |
| 85 | + |
| 86 | +#输出前10张图片及预测结果 |
| 87 | +k = 0 |
| 88 | +while k<10: |
| 89 | + #读取图像 |
| 90 | + print(X_test[k]) |
| 91 | + image = cv2.imread(X_test[k]) |
| 92 | + print(predictions_labels[k]) |
| 93 | + #显示图像 |
| 94 | + cv2.imshow("img", image) |
| 95 | + cv2.waitKey(0) |
| 96 | + cv2.destroyAllWindows() |
| 97 | + k = k + 1 |
0 commit comments