1，根据tensorflow -cifar10 示例 改进，以便适应更多图片与分类。2，完善打包图片到Bin文件的生成机制
python 3.5
numpy==1.12.0
tensorflow==1.2.1
第一步：下载训练数据集 （200多种图片分类，图片来源：ILSVRC2012，图片压缩过） 链接：https://pan.baidu.com/s/1lbM9-qs1BA2rzRFviZ3t6A （2021.12.02已更新） 提取码：59al
第二步：运行genar_train_data.py 其中训练集路径记得修改一下
第三步：开始训练 cifar10_train.py
cifar10_imagenet_eval.py 使用训练集评估模型效果
cifar10_runsingle.py 单独识别一张图片分类
详细解说文章：https://www.cnblogs.com/7rhythm/p/7091624.html 源代码与文章中代码片段已经有所不同，注意区分
Apache License
https://github.com/tensorflow/tensorflow/blob/master/LICENSE
- Improved according to the tensorflow-cifar10 example to accommodate more pictures and classifications.2. Improve the generation mechanism of packing pictures into Bin files
Python 3.5 Numpy = = 1.12.0 Tensorflow = = 1.2.1
Step 1: download the training data set (more than 200 image categories, image source: ILSVRC2012, image compressed) https://pan.baidu.com/s/1lbM9-qs1BA2rzRFviZ3t6A extracted code: 59al Step 2: run genar_train_data.py where path of the training set is modified Step 3: run the cifar10_train.py
Cifar10_imagenet_eval.Py The effect of the model was evaluated using the training set
Cifar10_runsingle.Py Identify a single image for classification
explanation article: https://www.cnblogs.com/7rhythm/p/7091624.html （source code and article code snippet is different, pay attention to distinguish）