Update readme.md

Unet/Pretreatment/readme.md

@@ -1,11 +1,43 @@
 # Little programs that preprocess raw data
 
-Contents:
-- 
+Contents: each program function is shown in the file name
+
+**Tips**:
+- Quickly open .ipynb file: Open the .ipynb file is very slow in Github and often doesn't load! So when you want to quickly check a .ipynb file from the Github, firstly you can copy the URL of this file. For example there is a .ipynb file in my Github repository, you can easily find this information: https://github.com/GaoBoYu599/Deep-Learning-Applications/blob/master/Unet/Pretreatment/全处理工程：一共4步.ipynb. Secondly, you copy this path into the https://nbviewer.jupyter.org/ , and the .ipynv file can be opened in this website. 
+
+**全过程处理：一共4步.ipynb**:
+- Set the parameters for preprocessing, including getting the size of the original image, setting the size of each sub-image, calculating the total number of sub-images
+- Cut the original image and save all sub-images
+- Turn all images' content into only 0 and 1 values and then resave them: these are the original dataset
+- Assemble all the sub-images in the original order into the original image
+
+**多图合并.ipynb**：
+- Select some adjacent sub-images for merging. For example, select 4 horizontal and 4 vertical sub-images in the upper left corner for regional splicing and merging.
+- This program can be used to 01 sub-images and original 3-channel sub-images
+- Save the assembled files in the same file format
+
+**subplot绘图调整.ipynb**:
+- Compare and contrast the label sub-images with the original sub-images
 
 ---
 
 # 各种处理原始数据的小程序
 
-内容：
--
+内容：各个程序的功能如文件名所示
+
+**注意**:
+- 快速打开.ipynb文件：因为在github里直接打开.ipynb文件非常缓慢甚至是加载不出来。所有如果你想快速查看一个github里的.ipynb文件，首先你先找到这个文件的网址（URL），比如在这个仓库下有一个.ipynb文件，可以很容易找到它的URL是：https://github.com/GaoBoYu599/Deep-Learning-Applications/blob/master/Unet/Pretreatment/全处理工程：一共4步.ipynb 。然后，直接拷贝这个URL到https://nbviewer.jupyter.org/ 中，就可以在这个网页下快速查看那个.ipynb文件。
+
+**全过程处理：一共4步.ipynb**:
+- 设置预处理的参数，包括获取原始图像的尺寸，设置分割子图的尺寸，计算总子图个数
+- 切割原始大图并保存所有子图
+- 将所有原始三通道彩色子图转换为只有0和1值的标签图：这些转换后的标签图就是原始的数据集
+- 将所有分割后子图按照原始顺序拼合为原来的大图
+
+**多图合并.ipynb**：
+- 选取部分彼此相邻的子图进行合并，例如：选取左上角横向4个纵向4个，总计共16个子图进行区域拼接合并。
+- 按照相同的文件格式保存区域合并后的图像
+
+**subplot绘图调整.ipynb**:
+- 专门负责绘制原始子图和标签子图的对比图
+