加载图像

    import tensorflow as tf
    import glob
    from itertools import groupby
    from collections import defaultdict
    sess = tf.InteractiveSession()
    image_filenames = glob.glob("./imagenet-dogs/n02*/*.jpg")
    image_filenames[0:2]
    training_dataset = defaultdict(list)
    testing_dataset = defaultdict(list)
    image_filename_with_breed = map(lambda filename: (filename.split("/")[2], filename), image_filenames)
    for dog_breed, breed_images in groupby(image_filename_with_breed, lambda x: x[0]):
        for i, breed_image in enumerate(breed_images):
            if i % 5 == 0:
                testing_dataset[dog_breed].append(breed_image[1])
            else:
                training_dataset[dog_breed].append(breed_image[1])
        breed_training_count = len(training_dataset[dog_breed])
        breed_testing_count = len(testing_dataset[dog_breed])
        breed_training_count_float = float(breed_training_count)
        breed_testing_count_float = float(breed_testing_count)
        assert round(breed_testing_count_float / (breed_training_count_float + breed_testing_count_float), 2) > 0.18, "Not enough testing images."
    print "training_dataset testing_dataset END ------------------------------------------------------"
    def write_records_file(dataset, record_location):
        writer = None
        current_index = 0
        for breed, images_filenames in dataset.items():
            for image_filename in images_filenames:
                if current_index % 100 == 0:
                    if writer:
                        writer.close()
                    record_filename = "{record_location}-{current_index}.tfrecords".format(
                        record_location=record_location,
                        current_index=current_index)
                    writer = tf.python_io.TFRecordWriter(record_filename)
                    print record_filename + "------------------------------------------------------" 
                current_index += 1
                image_file = tf.read_file(image_filename)
                try:
                    image = tf.image.decode_jpeg(image_file)
                except:
                    print(image_filename)
                    continue
                grayscale_image = tf.image.rgb_to_grayscale(image)
                resized_image = tf.image.resize_images(grayscale_image, [250, 151])
                image_bytes = sess.run(tf.cast(resized_image, tf.uint8)).tobytes()
                image_label = breed.encode("utf-8")
                example = tf.train.Example(features=tf.train.Features(feature={
                    'label': tf.train.Feature(bytes_list=tf.train.BytesList(value=[image_label])),
                    'image': tf.train.Feature(bytes_list=tf.train.BytesList(value=[image_bytes]))
                }))
                writer.write(example.SerializeToString())
        writer.close()
    write_records_file(testing_dataset, "./output/testing-images/testing-image")
    write_records_file(training_dataset, "./output/training-images/training-image")
    print "write_records_file testing_dataset training_dataset END------------------------------------------------------"
    filename_queue = tf.train.string_input_producer(
    tf.train.match_filenames_once("./output/training-images/*.tfrecords"))
    reader = tf.TFRecordReader()
    _, serialized = reader.read(filename_queue)
    features = tf.parse_single_example(
    serialized,
        features={
            'label': tf.FixedLenFeature([], tf.string),
            'image': tf.FixedLenFeature([], tf.string),
        })
    record_image = tf.decode_raw(features['image'], tf.uint8)
    image = tf.reshape(record_image, [250, 151, 1])
    label = tf.cast(features['label'], tf.string)
    min_after_dequeue = 10
    batch_size = 3
    capacity = min_after_dequeue + 3 * batch_size
    image_batch, label_batch = tf.train.shuffle_batch(
        [image, label], batch_size=batch_size, capacity=capacity, min_after_dequeue=min_after_dequeue)
    print "load image from TFRecord END------------------------------------------------------"
    float_image_batch = tf.image.convert_image_dtype(image_batch, tf.float32)
    conv2d_layer_one = tf.contrib.layers.convolution2d(
        float_image_batch,
        num_outputs=32,
        kernel_size=(5,5),
        activation_fn=tf.nn.relu,
        weights_initializer=tf.random_normal,
        stride=(2, 2),
        trainable=True)
    pool_layer_one = tf.nn.max_pool(conv2d_layer_one,
        ksize=[1, 2, 2, 1],
        strides=[1, 2, 2, 1],
        padding='SAME')
    conv2d_layer_one.get_shape(), pool_layer_one.get_shape()
    print "conv2d_layer_one pool_layer_one END------------------------------------------------------"
    conv2d_layer_two = tf.contrib.layers.convolution2d(
        pool_layer_one,
        num_outputs=64,
        kernel_size=(5,5),
        activation_fn=tf.nn.relu,
        weights_initializer=tf.random_normal,
        stride=(1, 1),
        trainable=True)
    pool_layer_two = tf.nn.max_pool(conv2d_layer_two,
        ksize=[1, 2, 2, 1],
        strides=[1, 2, 2, 1],
        padding='SAME')
    conv2d_layer_two.get_shape(), pool_layer_two.get_shape()
    print "conv2d_layer_two pool_layer_two END------------------------------------------------------"
    flattened_layer_two = tf.reshape(
        pool_layer_two,
        [
            batch_size,
            -1
        ])
    flattened_layer_two.get_shape()
    print "flattened_layer_two END------------------------------------------------------"
    hidden_layer_three = tf.contrib.layers.fully_connected(
        flattened_layer_two,
        512,
        weights_initializer=lambda i, dtype: tf.truncated_normal([38912, 512], stddev=0.1),
        activation_fn=tf.nn.relu
    )
    hidden_layer_three = tf.nn.dropout(hidden_layer_three, 0.1)
    final_fully_connected = tf.contrib.layers.fully_connected(
        hidden_layer_three,
        120,
        weights_initializer=lambda i, dtype: tf.truncated_normal([512, 120], stddev=0.1)
    )
    print "final_fully_connected END------------------------------------------------------"
    labels = list(map(lambda c: c.split("/")[-1], glob.glob("./imagenet-dogs/*")))
    train_labels = tf.map_fn(lambda l: tf.where(tf.equal(labels, l))[0,0:1][0], label_batch, dtype=tf.int64)
    loss = tf.reduce_mean(
        tf.nn.sparse_softmax_cross_entropy_with_logits(
            final_fully_connected, train_labels))
    batch = tf.Variable(0)
    learning_rate = tf.train.exponential_decay(
        0.01,
        batch * 3,
        120,
        0.95,
        staircase=True)
    optimizer = tf.train.AdamOptimizer(
        learning_rate, 0.9).minimize(
        loss, global_step=batch)
    train_prediction = tf.nn.softmax(final_fully_connected)
    print "train_prediction END------------------------------------------------------"
    filename_queue.close(cancel_pending_enqueues=True)
    coord.request_stop()
    coord.join(threads)
    print "END------------------------------------------------------"

学学笔记TF016:CNN达成、数据集、TFRecord、加载图像、模型、陶冶、调节和测试,tf01陆tfrecord

亚历克斯Net(AlexKrizhevsky,ILSV奥德赛C2013季军)适合做图像分类。层自左向右、自上向下读取,关联层分为一组,高度、宽度减小,深度扩张。深度扩充减弱互联网总结量。

磨练模型数据集 StanfordComputer视觉站点Stanford Dogs
http://vision.stanford.edu/aditya86/ImageNetDogs/
。数据下载解压到模型代码同一路线imagenet-dogs目录下。包蕴的120种狗图像。五分四教练,二成测试。产品模型必要预留原始数据交叉验证。每幅图像JPEG格式(奥德赛GB),尺寸不1。

图像转TFRecord文件,有助加快磨炼,简化图像标签相配,图像分离利用检查点文件对模型举行不间断测试。转变图像格式把颜色空间转灰度,图像修改统壹尺寸,标签除上每幅图像。磨练前只进行二次预处理,时间较长。

glob.glob
枚举内定路径目录,彰显数据集文件结构。“*”通配符能够完毕模糊查找。文件名中几个数字对应ImageNet体系WordNetID。ImageNet网址可用WordNetID查图像细节:
http://www.image-net.org/synset?wnid=n02085620

文件名分解为品种和相应的文件名,品种对应文件夹名称。依据品种对图像分组。枚举每一种品种图像,伍分一图像划入测试集。检查每一种门类测试图像是还是不是至少有方方面面图像的1八%。目录和图像协会到多少个与各样门类有关的字典,包涵各等级次序全体图像。分类图像协会到字典中,简化选拔分类图像及分类进程。

预管理阶段,依次遍历全部分类图像,展开列表中文件。用dataset图像填充TFRecord文件,把项目包蕴进去。dataset键值对应文件列表标签。record_location
存款和储蓄TFRecord输出路线。枚举dataset,当前目录用于文书划分,每隔十0m幅图像,训练样本音讯写入新的TFRecord文件,加速写操作进程。不大概被TensorFlow识别为JPEG图像,用try/catch忽略。转为灰度图减弱总括量和内存占用。tf.cast把奥迪Q伍GB值调换成[0,壹)区间内。标签按字符串存款和储蓄较高速,最棒调换为整数索引或独热编码秩一张量。

开发每幅图像,调换为灰度图,调节尺寸,增添到TFRecord文件。tf.image.resize_images函数把富有图像调解为一样尺寸,不思考长度宽度比,有扭动。裁剪、边界填充能保持图像长度宽度比。

鲁人持竿TFRecord文件读取图像,每一趟加载一丢丢图像及标签。修改图像形状有助演练和输出可视化。相配全部在教练集目录下TFRecord文件加载锻练图像。每种TFRecord文件包括多幅图像。tf.parse_single_example只从文件提取单个样本。批运算可同时磨练多幅图像或单幅图像,须要丰盛系统内部存储器。

图像转灰度值为[0,一)浮点类型,相称convolution2d希望输入。卷积输出第二维和最终1维不退换,中间两维产生变化。tf.contrib.layers.convolution2d开立模型第二层。weights_initializer设置正态随机值,第叁组滤波器填充正态布满随机数。滤波器设置trainable,新闻输入互联网,权值调节,进步模型正确率。
max_pool把出口降采集样品。ksize、strides
([1,2,2,1]),卷积输出形状减半。输出形状减小,不改动滤波器数量(输出通道)或图像批数量尺寸。减弱重量,与图像(滤波器)高度、宽度有关。愈来愈多输出通道,滤波器数量增添,贰倍于第二层。多少个卷积和池化层减弱输入中度、宽度,扩大吃水。诸多架构,卷积层和池化层超越伍层。练习调节和测试时间越来越长,能合营愈来愈多更纵横交错方式。
图像各种点与出口神经元创设全连接。softmax,全连接层要求二阶张量。第贰维区分图像,第3维输入张量秩一张量。tf.reshape
提示和行使此外全体维,-一把最终池化层调度为伟大秩一张量。
池化层展开,互连网当前景观与展望全连接层整合。weights_initializer接收可调用参数,lambda表达式重回截断正态布满,钦命布满标准差。dropout
削减模型中神经元主要性。tf.contrib.layers.fully_connected
输出前面全部层与教练中分类的全连接。各样像素与分类关联。互连网每一步将输入图像转化为滤波减小尺寸。滤波器与标签相称。减弱磨炼、测试网络计算量,输出更具一般性。

教练多少真实标签和模型预测结果,输入到教练优化器(优化每层权值)计算模型损失。多次迭代,每一遍升高模型准确率。大多数分拣函数(tf.nn.softmax)须要数值类型标签。每种标签调换代表包蕴全体分类列表索引整数。tf.map_fn
相称每一个标签并回到种类列表索引。map依附目录列表创制包罗分类列表。tf.map_fn
可用钦赐函数对数据流图张量映射,生成仅包括各类标签在具有类标签列表索引秩1张量。tf.nn.softmax用索引预测。

调解CNN,观察滤波器(卷积核)每轮迭代变化。设计精美CNN,第一个卷积层专门的学问,输入权值被自便起先化。权值通过图像激活,激活函数输出(特征图)随机。特征图可视化,输出外观与原始图相似,被施加静力(static)。静力由全数权值的放肆激发。经过多轮迭代,权值被调度拟合训练反馈,滤波器趋于1致。网络未有,滤波器与图像区别细小情势类似。tf.image_summary得报到并且接受集练习后的滤波器和个性图简单视图。数据流图图像概要输出(image
summary
output)从总体理解所选拔的滤波器和输入图像特点图。TensorDebugger,迭代中以GIF动画查看滤波器变化。

文件输入存款和储蓄在SparseTensor,半数以上轻重为0。CNN使用稠密输入,每一种值都主要,输入超过伍分之3份额非0。

 

    import tensorflow as tf
    import glob
    from itertools import groupby
    from collections import defaultdict
    sess = tf.InteractiveSession()
    image_filenames = glob.glob("./imagenet-dogs/n02*/*.jpg")
    image_filenames[0:2]
    training_dataset = defaultdict(list)
    testing_dataset = defaultdict(list)
    image_filename_with_breed = map(lambda filename: (filename.split("/")[2], filename), image_filenames)
    for dog_breed, breed_images in groupby(image_filename_with_breed, lambda x: x[0]):
        for i, breed_image in enumerate(breed_images):
            if i % 5 == 0:
                testing_dataset[dog_breed].append(breed_image[1])
            else:
                training_dataset[dog_breed].append(breed_image[1])
        breed_training_count = len(training_dataset[dog_breed])
        breed_testing_count = len(testing_dataset[dog_breed])
        breed_training_count_float = float(breed_training_count)
        breed_testing_count_float = float(breed_testing_count)
        assert round(breed_testing_count_float / (breed_training_count_float + breed_testing_count_float), 2) > 0.18, "Not enough testing images."
    print "training_dataset testing_dataset END ------------------------------------------------------"
    def write_records_file(dataset, record_location):
        writer = None
        current_index = 0
        for breed, images_filenames in dataset.items():
            for image_filename in images_filenames:
                if current_index % 100 == 0:
                    if writer:
                        writer.close()
                    record_filename = "{record_location}-{current_index}.tfrecords".format(
                        record_location=record_location,
                        current_index=current_index)
                    writer = tf.python_io.TFRecordWriter(record_filename)
                    print record_filename + "------------------------------------------------------" 
                current_index += 1
                image_file = tf.read_file(image_filename)
                try:
                    image = tf.image.decode_jpeg(image_file)
                except:
                    print(image_filename)
                    continue
                grayscale_image = tf.image.rgb_to_grayscale(image)
                resized_image = tf.image.resize_images(grayscale_image, [250, 151])
                image_bytes = sess.run(tf.cast(resized_image, tf.uint8)).tobytes()
                image_label = breed.encode("utf-8")
                example = tf.train.Example(features=tf.train.Features(feature={
                    'label': tf.train.Feature(bytes_list=tf.train.BytesList(value=[image_label])),
                    'image': tf.train.Feature(bytes_list=tf.train.BytesList(value=[image_bytes]))
                }))
                writer.write(example.SerializeToString())
        writer.close()
    write_records_file(testing_dataset, "./output/testing-images/testing-image")
    write_records_file(training_dataset, "./output/training-images/training-image")
    print "write_records_file testing_dataset training_dataset END------------------------------------------------------"
    filename_queue = tf.train.string_input_producer(
    tf.train.match_filenames_once("./output/training-images/*.tfrecords"))
    reader = tf.TFRecordReader()
    _, serialized = reader.read(filename_queue)
    features = tf.parse_single_example(
    serialized,
        features={
            'label': tf.FixedLenFeature([], tf.string),
            'image': tf.FixedLenFeature([], tf.string),
        })
    record_image = tf.decode_raw(features['image'], tf.uint8)
    image = tf.reshape(record_image, [250, 151, 1])
    label = tf.cast(features['label'], tf.string)
    min_after_dequeue = 10
    batch_size = 3
    capacity = min_after_dequeue + 3 * batch_size
    image_batch, label_batch = tf.train.shuffle_batch(
        [image, label], batch_size=batch_size, capacity=capacity, min_after_dequeue=min_after_dequeue)
    print "load image from TFRecord END------------------------------------------------------"
    float_image_batch = tf.image.convert_image_dtype(image_batch, tf.float32)
    conv2d_layer_one = tf.contrib.layers.convolution2d(
        float_image_batch,
        num_outputs=32,
        kernel_size=(5,5),
        activation_fn=tf.nn.relu,
        weights_initializer=tf.random_normal,
        stride=(2, 2),
        trainable=True)
    pool_layer_one = tf.nn.max_pool(conv2d_layer_one,
        ksize=[1, 2, 2, 1],
        strides=[1, 2, 2, 1],
        padding='SAME')
    conv2d_layer_one.get_shape(), pool_layer_one.get_shape()
    print "conv2d_layer_one pool_layer_one END------------------------------------------------------"
    conv2d_layer_two = tf.contrib.layers.convolution2d(
        pool_layer_one,
        num_outputs=64,
        kernel_size=(5,5),
        activation_fn=tf.nn.relu,
        weights_initializer=tf.random_normal,
        stride=(1, 1),
        trainable=True)
    pool_layer_two = tf.nn.max_pool(conv2d_layer_two,
        ksize=[1, 2, 2, 1],
        strides=[1, 2, 2, 1],
        padding='SAME')
    conv2d_layer_two.get_shape(), pool_layer_two.get_shape()
    print "conv2d_layer_two pool_layer_two END------------------------------------------------------"
    flattened_layer_two = tf.reshape(
        pool_layer_two,
        [
            batch_size,
            -1
        ])
    flattened_layer_two.get_shape()
    print "flattened_layer_two END------------------------------------------------------"
    hidden_layer_three = tf.contrib.layers.fully_connected(
        flattened_layer_two,
        512,
        weights_initializer=lambda i, dtype: tf.truncated_normal([38912, 512], stddev=0.1),
        activation_fn=tf.nn.relu
    )
    hidden_layer_three = tf.nn.dropout(hidden_layer_three, 0.1)
    final_fully_connected = tf.contrib.layers.fully_connected(
        hidden_layer_three,
        120,
        weights_initializer=lambda i, dtype: tf.truncated_normal([512, 120], stddev=0.1)
    )
    print "final_fully_connected END------------------------------------------------------"
    labels = list(map(lambda c: c.split("/")[-1], glob.glob("./imagenet-dogs/*")))
    train_labels = tf.map_fn(lambda l: tf.where(tf.equal(labels, l))[0,0:1][0], label_batch, dtype=tf.int64)
    loss = tf.reduce_mean(
        tf.nn.sparse_softmax_cross_entropy_with_logits(
            final_fully_connected, train_labels))
    batch = tf.Variable(0)
    learning_rate = tf.train.exponential_decay(
        0.01,
        batch * 3,
        120,
        0.95,
        staircase=True)
    optimizer = tf.train.AdamOptimizer(
        learning_rate, 0.9).minimize(
        loss, global_step=batch)
    train_prediction = tf.nn.softmax(final_fully_connected)
    print "train_prediction END------------------------------------------------------"
    filename_queue.close(cancel_pending_enqueues=True)
    coord.request_stop()
    coord.join(threads)
    print "END------------------------------------------------------"

 

参考资料:
《面向机器智能的TensorFlow实践》

接待加小编微信调换:qingxingfengzi
自个儿的微信公众号:qingxingfengzigz
本身老伴张幸清的微信公众号:qingqingfeifangz

http://www.bkjia.com/Pythonjc/1213552.htmlwww.bkjia.comtruehttp://www.bkjia.com/Pythonjc/1213552.htmlTechArticle学习笔记TF016:CNN实现、数据集、TFRecord、加载图像、模型、训练、调试,tf016tfrecord
亚历克斯Net(亚历克斯 Krizhevsky,ILSV牧马人C二〇一三季军)适合做图像分类。层自左…

图像转灰度值为[0,1)浮点类型,相称convolution2d愿意输入。卷积输出第2维和最终一维不变,中间两维发生变化。tf.contrib.layers.convolution二d开立模型第二层。weights_initializer设置正态随机值,第1组滤波器填充正态分布随机数。滤波器设置trainable,消息输入互连网,权值调度,升高模型正确率。
max_pool把出口降采集样品。ksize、strides
([1,2,2,1]),卷积输出形状减半。输出形状减小,不转移滤波器数量(输出通道)或图像批数量尺寸。收缩重量,与图像(滤波器)中度、宽度有关。越多输出通道,滤波器数量增多,二倍于第叁层。多少个卷积和池化层缩小输入中度、宽度,扩展吃水。繁多架构,卷积层和池化层超过伍层。演习调节和测试时间更长,能相配更加多更盘根错节形式。
图像每一种点与输出神经元创设全连接。softmax,全连接层须求二阶张量。第一维区分图像,第一维输入张量秩1张量。tf.reshape
提示和平运动用别的全体维,-壹把最后池化层调度为大侠秩一张量。
池化层张开,互连网当前情形与猜想全连接层整合。weights_initializer接收可调用参数,lambda表明式重返截断正态分布,钦赐布满规范差。dropout
削减模型中神经元主要性。tf.contrib.layers.fully_connected
输出前面全数层与练习中分类的全连接。每一个像素与分类关联。网络每一步将输入图像转化为滤波减小尺码。滤波器与标签相配。减少演练、测试互连网计算量,输出更具一般性。

参考资料:
《面向机器智能的TensorFlow施行》

亚历克斯Net(AlexKrizhevsky,ILSV中华VC二〇一二亚军)适合做图像分类。层自左向右、自上向下读取,关联层分为一组,低度、宽度减小,深度扩大。深度增添收缩网络总括量。

开辟每幅图像,调换为灰度图,调整尺寸,增多到TFRecord文件。tf.image.resize_images函数把全体图像调节为同1尺寸,不思虑长度宽度比,有扭动。裁剪、边界填充能保持图像长度宽度比。

锻炼模型数据集 StanfordComputer视觉站点Stanford Dogs
http://vision.stanford.edu/aditya86/ImageNetDogs/
。数据下载解压到模型代码同一路线imagenet-dogs目录下。包蕴的120种狗图像。八成演习,二成测试。产品模型要求预留原始数据交叉验证。每幅图像JPEG格式(宝马7系GB),尺寸不一。

图像转TFRecord文件,有助加快练习,简化图像标签相称,图像分离利用检查点文件对模型实行不间断测试。调换图像格式把颜色空间转灰度,图像修改统一尺寸,标签除上每幅图像。演练前只实行一回预管理,时间较长。

文本输入存款和储蓄在SparseTensor,大多数重量为0。CNN使用稠密输入,每一种值都主要,输入大多数分量非0。

教练多少真实标签和模型预测结果,输入到教练优化器(优化每层权值)总结模型损失。数十次迭代,每一回升高模型准确率。超越二分一分类函数(tf.nn.softmax)供给数值类型标签。每一种标签调换代表包括全数分类列表索引整数。tf.map_fn
相配种种标签并回到体体系表索引。map凭借目录列表创造包涵分类列表。tf.map_fn
可用钦定函数对数码流图张量映射,生成仅蕴涵各样标签在全数类标签列表索引秩一张量。tf.nn.softmax用索引预测。

glob.glob
枚举内定路径目录,显示数据集文件结构。“*”通配符能够达成模糊查找。文件名中七个数字对应ImageNet连串WordNetID。ImageNet网址可用WordNetID查图像细节:
http://www.image-net.org/synset?wnid=n02085620

预管理阶段,依次遍历全部分类图像,展开列表中文件。用dataset图像填充TFRecord文件,把品种包涵进去。dataset键值对应文件列表标签。record_location
存款和储蓄TFRecord输出路线。枚举dataset,当前目录用于文书划分,每隔十0m幅图像,演习样本新闻写入新的TFRecord文件,加快写操作进度。不可能被TensorFlow识别为JPEG图像,用try/catch忽略。转为灰度图减弱总结量和内存占用。tf.cast把索罗德GB值调换来[0,1)区间内。标签按字符串存款和储蓄较高速,最棒调换为整数索引或独热编码秩一张量。

安分守纪TFRecord文件读取图像,每一趟加载少许图像及标签。修改图像形状有助磨练和输出可视化。相配全部在教练集目录下TFRecord文件加载陶冶图像。每种TFRecord文件包罗多幅图像。tf.parse_single_example只从文件提取单个样本。批运算可同时磨炼多幅图像或单幅图像,须求丰硕系统内部存款和储蓄器。

调养CNN,观望滤波器(卷积核)每轮迭代变化。设计能够CNN,第三个卷积层职业,输入权值被狂妄开首化。权值通过图像激活,激活函数输出(特征图)随机。特征图可视化,输出外观与原始图相似,被施加静力(static)。静力由全部权值的轻便激发。经过多轮迭代,权值被调度拟合练习反馈,滤波器趋于同一。互联网未有,滤波器与图像差异细小形式类似。tf.image_summary得报到并且接受集陶冶后的滤波器和特征图轻易视图。数据流图图像概要输出(image
summary
output)从全部领悟所选用的滤波器和输入图像特点图。TensorDebugger,迭代中以GIF动画查看滤波器变化。

文本名分解为项目和呼应的文本名,品种对应文件夹名称。依靠品种对图像分组。枚举各种种类图像,十分二图像划入测试集。检查各类体系测试图像是还是不是至少有整套图像的1八%。目录和图像协会到八个与各样项目有关的字典,包罗各档期的顺序全部图像。分类图像组织到字典中,简化选取分类图像及分类进程。

 

接待加笔者微信调换:qingxingfengzi
本身的微信公众号:qingxingfengzigz
本身爱妻张幸清的微信公众号:qingqingfeifangz