写一百题程序学Python 【 11-20】

这次我试着利用python对qq群聊接收的图片进行了分类，还尝试涉及了下tensorflow（尽管目前我还是没想到哪里能用上）

11.int和string数据类型转换

t = input()
r = input()
res = int(t)*int(r)
print(res)

stringA = str(res)
stringA += " is worth"
print(stringA)

12-14  是通过一系列函数，来处理qq群聊收到的图片

12.读取文件夹下所有图片并获取宽度和高度，删除小于400pixel的图片

from PIL import Image
import os
path = "/media/Image/Group/test/"
for file in os.listdir(path):
    fp = open(path+"/"+file, 'rb')
    img = Image.open(fp)
    fp.close()
    x,y = img.size
    print(x,y)
    if x<400 or y<400:
          os.remove(path+"/"+file)
 

file变量只会保存文件名称但是并不会保存完整路径，之前没了解清楚结果以为是文件有特殊字符

13.基于12，追加判断是文件还是文件夹，如果是文件夹则不处理

from PIL import Image
import os
path = "/media/Image/Group"
i = 0
for file in os.listdir(path):
     if os.path.isdir(path+"/"+file) :
         break
     else:
        fp = open(path+"/"+file, 'rb')
        img = Image.open(fp)
        fp.close()
        x,y = img.size
        print(x,y)
        if x<400 or y<400:
            os.remove(path+"/"+file)
            i = i+1
print("Clear "+str(i)+" images")

14.基于13，追加判断图片的主要色彩，主要用于删除掉网页截图和二维码还有qq截图

扫描二维码关注公众号，回复： 2797923 查看本文章

主要步骤是，先尝试转图片为HSV，然后按照比重，获取最主要的颜色，依照颜色是否一致（说明是纯色，可能是背景），颜色是否低于高于一定值（同样比较可能是背景），来择选图片

当然在主要步骤之前，依然会进行图片大小的筛选

from PIL import Image
import os
import shutil
import colorsys
path = "/media/Image/Group"
path1 = "/media/Image/Group/Image1"
path2 = "/media/Image/Group/Image2"
path3 = "/media/Image/Group/Image3"
samepath = "/media/Image/Group/xysame"
testpath = "/media/Image/Group/test"
notpath = "/media/Image/Group/bedelete/"
def qqimage_getcolor(filepath):
    image = Image.open(filepath)
    image.thumbnail((512,512))
    max_score = 0.0001
    dominant_color = None
    for count, (r, g, b) in image.getcolors(image.size[0] * image.size[1]):
        saturation = colorsys.rgb_to_hsv(r / 255.0, g / 255.0, b / 255.0)[1]
        y = min(abs(r * 2104 + g * 4130 + b * 802 + 4096 + 131072) >> 13, 235)
        y = (y - 16.0) / (235 - 16)
        if y > 0.9:
            continue
        score = (saturation + 0.1) * count
        if score > max_score:
            max_score = score
            dominant_color = (r, g, b)
    return dominant_color

def qqimage_move(path1,path2):
    shutil.copy(path1, path2)
    os.remove(path1)

def qqimage_processer(path,xypath,recyclepath):
    i = 0
    j = 0
    for file in os.listdir(path):
         if os.path.isdir(path+"/"+file) :
             break
         else:
            fp = open(path+"/"+file, 'rb')
            img = Image.open(fp)
            fp.close()
            x,y = img.size
            #print(x,y)
            if x<400 or y<400:
                qqimage_move(path + "/" + file, recyclepath)
                i = i+1
            elif x == y:
                j = j+1
                qqimage_move(path+"/"+file,xypath)
            else:
                try:
                    color = qqimage_getcolor(path+"/"+file)
                    print(file)
                    print(color)
                    if color[0] == color[1] and color[1] == color[2] and (color[0] <= 10 or color[0] >= 240):
                        qqimage_move(path + "/" + file, recyclepath)
                        i = i + 1
                except:
                    print("image not support")
    print("+++++++++++++++++++++++++++++++++++++++++")
    print("Clear "+str(i)+" images which smaller than 400 pixel")
    print("Have "+str(j)+" height and width are same")
    print("+++++++++++++++++++++++++++++++++++++++++")



qqimage_processer(path,samepath,notpath)

参考：

http://www.jb51.net/article/68958.htm   python获取文件扩展名的方法

https://blog.csdn.net/lhh31/article/details/52015500 python图像主色调（主颜色）提取

http://www.jb51.net/article/62526.htm   Python通过PIL获取图片主要颜色并和颜色库进行对比的方法

15.使用tensorflow运算矩阵

           

import os
import numpy as np
import tensorflow as tf

#一对括号表示的是列表，两对括号才表示矩阵
matrix1 = tf.constant([[3,3]])
matrix2 = tf.constant([[2],
                       [2]])#这个不太懂为什么要这样写
product = tf.matmul(matrix1,matrix2)


sess= tf.Session()
result = sess.run(product)
print(result)
sess.close()

参考：https://morvanzhou.github.io/tutorials/machine-learning/tensorflow/2-4-variable/   

16.tensorflow中实现 x=x+1

import os
import numpy as np
import tensorflow as tf

#set a variable
state = tf.Variable(0, name='counter')

 #set a constant
one = tf.constant(1)

 #set a algorithm with tensorflow
new_value = tf.add(state,one)

 #make value save to "state" variable
update = tf.assign(state,new_value)

#init all variable
init = tf.global_variables_initializer()

 #run session
with tf.Session() as sess:
    sess.run(init)
    for _ in range(10):
        sess.run(update)
        print(sess.run(state))

17.tensorflow外部赋值

主要应用placeholder函数，使用这个函数可以定义支持外部传值的变量，配合feed_dict= 传值

import os
import numpy as np
import tensorflow as tf

input1 = tf.placeholder(tf.float32)
input2 = tf.placeholder(tf.float32)

output = tf.multiply(input1,input2)

i = 10
j = 20
with tf.Session() as sess:
     print(sess.run(output,feed_dict={input1:[7,8,10,i],input2:[2,4,11,j]}))

18.python按照顺序批量重命名文件

def serialrenamed(path1,path2):
    i = 0
    for file in os.listdir(path1):
        if path1 + "/" + file != path2 + "/" + str(i) + os.path.splitext(file)[1]: #如果名称不重复才进行处理，否则跳过
             shutil.copy(path1 + "/" + file,path2 + "/" + str(i) + os.path.splitext(file)[1] )
             os.remove(path1+ "/" + file)
             i = i + 1

19.创建等差数列

import numpy as np
group = np.linspace(0,100,50)
print(group)

如果在linespace函数后加上[:, np.newaxis]  ，则每次生成等差数存储时都会创建新的轴（列）

import numpy as np
group = np.linspace(0,100,50)[:, np.newaxis]
print(group)

20.产生正态分布的随机数

产生100个

import numpy as np

s = np.random.normal(loc = 0,scale = 1.0, size=100)

print(s)

产生1个

import numpy as np

s = np.random.normal(loc = 0,scale = 1.0)

print(s)