博客
关于我
Keras自定义网络进行十分类图像识别
阅读量:262 次
发布时间:2019-03-01

本文共 4177 字,大约阅读时间需要 13 分钟。

import osimport numpy as npimport tensorflow as tfimport randomimport seaborn as snsimport matplotlib.pyplot as pltfrom keras.models import Sequential, Modelfrom keras.layers import Dense, Dropout, Activation, Flatten, Inputfrom keras.layers.convolutional import Conv2D, MaxPooling2Dfrom keras.optimizers import RMSprop, Adam, SGDfrom keras.preprocessing import imagefrom keras.preprocessing.image import ImageDataGeneratorfrom keras.utils import np_utilsfrom sklearn.model_selection import train_test_split

图片预处理

def read_and_process_image(data_dir,width=32, height=32, channels=3, preprocess=False):        train_classes= [data_dir +  i for i in os.listdir(data_dir) ]    train_images = []    for train_class in train_classes:        train_images= train_images + [train_class + "/" + i for i in os.listdir(train_class)]        random.shuffle(train_images)        def read_image(file_path, preprocess):        img = image.load_img(file_path, target_size=(height, width))        x = image.img_to_array(img)        x = np.expand_dims(x, axis=0)        # if preprocess:            # x = preprocess_input(x)        return x        def prep_data(images, proprocess):        count = len(images)        data = np.ndarray((count, height, width, channels), dtype = np.float32)                for i, image_file in enumerate(images):            image = read_image(image_file, preprocess)            data[i] = image                return data        def read_labels(file_path):        labels = []        for i in file_path:            if 'airplane' in i:                label = 0            elif 'automobile' in i:                label = 1            elif 'bird' in i:                label = 2            elif 'cat' in i:                label = 3            elif 'deer' in i:                label = 4            elif 'dog' in i:                label = 5            elif 'frog' in i:                label = 6            elif 'horse' in i:                label = 7            elif 'ship' in i:                label = 8            elif 'truck' in i:                label = 9            labels.append(label)                return labels        X = prep_data(train_images, preprocess)    labels = read_labels(train_images)        assert X.shape[0] == len(labels)        print("Train shape: {}".format(X.shape))        return X, labels

读取训练集,以及测试集

# 读取训练集图片WIDTH = 32HEIGHT = 32CHANNELS = 3X, y = read_and_process_image('D:/Python Project/cifar-10/train/',width=WIDTH, height=HEIGHT, channels=CHANNELS)# 读取测试集图片WIDTH = 32HEIGHT = 32CHANNELS = 3test_X, test_y = read_and_process_image('D:/Python Project/cifar-10/test/',width=WIDTH, height=HEIGHT, channels=CHANNELS)# 统计ysns.countplot(y)# 统计ysns.countplot(test_y)

one-hot编码

train_y = np_utils.to_categorical(y)test_y = np_utils.to_categorical(test_y)

显示图片

# 显示图片def show_picture(X, idx):    plt.figure(figsize=(10,5), frameon=True)    img = X[idx,:,:,::-1]    img = img/255    plt.imshow(img)    plt.show()for idx in range(0,3):    show_picture(X, idx)

定义模型

num_classes=10model = Sequential()model.add(Conv2D(32 ,3 ,input_shape=(HEIGHT,WIDTH,CHANNELS),activation='relu',padding='same'))model.add(Conv2D(32 ,3 ,activation='relu',padding='same'))model.add(MaxPooling2D(pool_size=2))model.add(Conv2D(64 ,3 ,activation='relu',padding='same'))model.add(Conv2D(64 ,3 ,activation='relu',padding='same'))model.add(MaxPooling2D(pool_size=2))model.add(Conv2D(128 ,3 ,activation='relu',padding='same'))model.add(Conv2D(128 ,3 ,activation='relu',padding='same'))model.add(MaxPooling2D(pool_size=2))model.add(Conv2D(256 ,3 ,activation='relu',padding='same'))model.add(Conv2D(256 ,3 ,activation='relu',padding='same'))model.add(MaxPooling2D(pool_size=2))model.add(Flatten())model.add(Dense(256, activation='relu'))model.add(Dropout(0.5))model.add(Dense(256, activation='relu'))model.add(Dropout(0.5))model.add(Dense(num_classes, activation='softmax'))model.compile(loss='categorical_crossentropy',optimizer='adam',metrics=['accuracy'])model.summary()

训练模型

history = model.fit(X,train_y, validation_data=(test_X, test_y),epochs=20,batch_size=100,verbose=True)score = model.evaluate(test_X, test_y, verbose=0)print("Large CNN Error: %.2f%%" %(100-score[1]*100))

 

转载地址:http://kshv.baihongyu.com/

你可能感兴趣的文章
Mysql学习总结(42)——MySql常用脚本大全
查看>>
Mysql学习总结(43)——MySQL主从复制详细配置
查看>>
Mysql学习总结(44)——Linux下如何实现mysql数据库每天自动备份定时备份
查看>>
Mysql学习总结(45)——Mysql视图和事务
查看>>
Mysql学习总结(46)——8种常被忽视的SQL错误用法
查看>>
Mysql学习总结(48)——MySql的日志与备份还原
查看>>
Mysql学习总结(49)——从开发规范、选型、拆分到减压
查看>>
Mysql学习总结(4)——MySql基础知识、存储引擎与常用数据类型
查看>>
Mysql学习总结(50)——Oracle,mysql和SQL Server的区别
查看>>
Mysql学习总结(51)——Linux主机Mysql数据库自动备份
查看>>
Mysql学习总结(52)——最全面的MySQL 索引详解
查看>>
Mysql学习总结(53)——使用MySql开发的Java开发者规范
查看>>
Mysql学习总结(54)——MySQL 集群常用的几种高可用架构方案
查看>>
Mysql学习总结(55)——MySQL 语句大全再温习
查看>>
Mysql学习总结(56)——MySQL用户管理和权限设置
查看>>
Mysql学习总结(57)——MySQL查询当天、本周、本月、上周、本周、上月、距离当前现在6个月数据
查看>>
Mysql学习总结(58)——深入理解Mysql的四种隔离级别
查看>>
Mysql学习总结(59)——数据库分库分表策略总结
查看>>
Mysql学习总结(5)——MySql常用函数大全讲解
查看>>
Mysql学习总结(60)——并发量大、数据量大的互联网业务数据库设计规范总结
查看>>