效果图

相似度/字符

识别的图片：

代码

from PIL import Image
import hashlib
import time
import os
import math

class VectorCompare: # 向量比较类
    def magnitude(self,concordance): # 计算向量长度
        total = 0
        for word,count in concordance.items():
            total += count ** 2
        return math.sqrt(total)

    def relation(self,concordance1, concordance2):#求两个向量的关系值？协。
        relevance = 0
        topvalue = 0
        for word, count in concordance1.items():
            if word in concordance2:
                topvalue += count * concordance2[word]
        return topvalue / (self.magnitude(concordance1) * self.magnitude(concordance2))



def buildvector(im): # 创建向量
    d1 = {}

    count = 0
    for i in im.getdata():
        d1[count] = i
        count += 1 #计数值

    return d1
# 返回类型是一个字典
v = VectorCompare()
# 图标集
iconset =  ['0','1','2','3','4','5','6','7','8','9','0','a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z']
# 图片集
imageset = []

for letter in iconset:
    for img in os.listdir('./iconset/%s/'%(letter)):
        temp = []
        if img != "Thumbs.db" and img != ".DS_Store": # 找到目录inconset下的样例图片
            temp.append(buildvector(Image.open("./iconset/%s/%s"%(letter,img)))) #将图片转化为向量存给temp
        imageset.append({letter:temp})  # 将图片集合加入 字母和样例的集合


im = Image.open("captcha.gif") #打开我们要检测的图片
im2 = Image.new("P",im.size,255) #使用p模式打开，色彩255
im.convert("P") # 将im的模式转换位P
temp = {} #暂存空间

for x in range(im.size[1]): # 宽度
    for y in range(im.size[0]): # 高度
        pix = im.getpixel((y,x)) #获取像素存到字典中
        temp[pix] = pix # ？
        if pix == 220 or pix == 227: # these are the numbers to get
            im2.putpixel((y,x),0) # im2导入

inletter = False
foundletter=False
start = 0
end = 0

letters = []

for y in range(im2.size[0]): # slice across
    for x in range(im2.size[1]): # slice down
        pix = im2.getpixel((y,x))
        if pix != 255:
            inletter = True
            break

    if foundletter == False and inletter == True:
        foundletter = True
        start = y

    if foundletter == True and inletter == False:
        foundletter = False
        end = y
        letters.append((start,end))


    inletter=False

count = 0
for letter in letters:
    im3 = im2.crop(( letter[0] , 0, letter[1],im2.size[1] ))

    guess = []

    for image in imageset:
        for x,y in image.items():
            if len(y) != 0:
                guess.append( ( v.relation(y[0],buildvector(im3)),x) )

    guess.sort(reverse=True)
    print (guess[0])
    count += 1