强智教务系统验证码识别 OpenCV
强智教务系统验证码识别 OpenCV
强智教务系统验证码验证码字符位置相对固定，比较好切割
找准切割位置，将其分为四部分，匹配自建库即可，识别率近乎100%，如果觉得不错，点个star吧
https://github.com/WindrunnerMax/SWVerifyCode 提供Java、PHP、Python、JavaScript版本
首先使用代码切割验证码，挑选出切割的比较好的验证码，制作比对库
由于使用matchTemplate函数，要求待匹配图必须比库图小，于是需要放大库图边界
TestImgCut.py切割图片并挑选合适切割位置
#!/usr/bin/python
# -*- coding: utf-8 -*-
from fnmatch import fnmatch
from queue import Queue
import matplotlib.pyplot as plt
import cv2
import time
import os
from Convert import Convert
import requests
def _get_static_binary_image(img, threshold = 140):
'''
手动二值化
'''
img = Image.open(img)
img = img.convert('L')
pixdata = img.load()
w, h = img.size
for y in range(h):
for x in range(w):
if pixdata[x, y] < threshold:
pixdata[x, y] = 0
else:
pixdata[x, y] = 255
return img
def cfs(im,x_fd,y_fd):
'''用队列和集合记录遍历过的像素坐标代替单纯递归以解决cfs访问过深问题
'''
# print('**********')
xaxis=[]
yaxis=[]
visited =set()
q = Queue()
q.put((x_fd, y_fd))
visited.add((x_fd, y_fd))
offsets=[(1, 0), (0, 1), (-1, 0), (0, -1)]#四邻域
while not q.empty():
x,y=q.get()
for xoffset,yoffset in offsets:
x_neighbor,y_neighbor = x+xoffset,y+yoffset
if (x_neighbor,y_neighbor) in (visited):
continue # 已经访问过了
visited.add((x_neighbor, y_neighbor))
try:
if im[x_neighbor, y_neighbor] == 0:
xaxis.append(x_neighbor)
yaxis.append(y_neighbor)
q.put((x_neighbor,y_neighbor))
except IndexError:
pass
# print(xaxis)
if (len(xaxis) == 0 | len(yaxis) == 0):
xmax = x_fd + 1
xmin = x_fd
ymax = y_fd + 1
ymin = y_fd
else:
xmax = max(xaxis)
xmin = min(xaxis)
ymax = max(yaxis)
ymin = min(yaxis)
#ymin,ymax=sort(yaxis)
return ymax,ymin,xmax,xmin
def detectFgPix(im,xmax):
'''搜索区块起点
'''
h,w = im.shape[:2]
for y_fd in range(xmax+1,w):
for x_fd in range(h):
if im[x_fd,y_fd] == 0:
return x_fd,y_fd
def CFS(im):
'''切割字符位置
'''
zoneL=[]#各区块长度L列表
zoneWB=[]#各区块的X轴[起始，终点]列表
zoneHB=[]#各区块的Y轴[起始，终点]列表
xmax=0#上一区块结束黑点横坐标,这里是初始化
for i in range(10):
try:
x_fd,y_fd = detectFgPix(im,xmax)
# print(y_fd,x_fd)
xmax,xmin,ymax,ymin=cfs(im,x_fd,y_fd)
L = xmax - xmin
H = ymax - ymin
zoneL.append(L)
zoneWB.append([xmin,xmax])
zoneHB.append([ymin,ymax])
except TypeError:
return zoneL,zoneWB,zoneHB
return zoneL,zoneWB,zoneHB
def cutting_img(im,im_position,xoffset = 1,yoffset = 1):
# 识别出的字符个数
im_number = len(im_position[1])
if(im_number>=4): im_number = 4;
imgArr = []
# 切割字符
for i in range(im_number):
im_start_X = im_position[1][i][0] - xoffset
im_end_X = im_position[1][i][1] + xoffset
im_start_Y = im_position[2][i][0] - yoffset
im_end_Y = im_position[2][i][1] + yoffset
cropped = im[im_start_Y:im_end_Y, im_start_X:im_end_X]
imgArr.append(cropped)
cv2.imwrite(str(i)+"v.jpg",cropped) # 查看切割效果
return im_number,imgArr
def main():
cvt = Convert()
req = requests.get("http://XXXXXXXXXXXXXXXXXXX/verifycode.servlet")
img = cvt.run(req.content)
cv2.imwrite("v.jpg",img)
#切割的位置
im_position = CFS(img) # Auto
print(im_position)
maxL = max(im_position[0])
minL = min(im_position[0])
# 如果有粘连字符，如果一个字符的长度过长就认为是粘连字符，并从中间进行切割
if(maxL > minL + minL * 0.7):
maxL_index = im_position[0].index(maxL)
minL_index = im_position[0].index(minL)
# 设置字符的宽度
im_position[0][maxL_index] = maxL // 2
im_position[0].insert(maxL_index + 1, maxL // 2)
# 设置字符X轴[起始，终点]位置
im_position[1][maxL_index][1] = im_position[1][maxL_index][0] + maxL // 2
im_position[1].insert(maxL_index + 1, [im_position[1][maxL_index][1] + 1, im_position[1][maxL_index][1] + 1 + maxL // 2])
# 设置字符的Y轴[起始，终点]位置
im_position[2].insert(maxL_index + 1, im_position[2][maxL_index])
# 切割字符，要想切得好就得配置参数，通常 1 or 2 就可以
cutting_img_num,imgArr = cutting_img(img,im_position,1,1)
# # 直接使用库读取图片识别验证码
# result=""
# for i in range(cutting_img_num):
# try:
# template = imgArr[i]
# tempResult=""
# matchingDegree=0.0
# filedirWarehouse = '../../Warehouse/StrIntell/'
# for fileImg in os.listdir(filedirWarehouse):
# if fnmatch(fileImg, '*.jpg'):
# # print(file)
# img = cv2.imread(filedirWarehouse+fileImg,0)
# res = cv2.matchTemplate(img,template,3) #img原图 template模板 用模板匹配原图
# min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res)
# # print(str(i)+" "+file.split('.')[0]+" "+str(max_val))
# if(max_val>matchingDegree):
# tempResult=fileImg.split('.')[0]
# matchingDegree=max_val
# result+=tempResult
# matchingDegree=0.0
# except Exception as err:
# print("ERROR "+ str(err))
# pass
# print('切图：%s' % cutting_img_num)
# print('识别为：%s' % result)
if __name__ == '__main__':
main()
resize.py改变图片边界
import cv2
from fnmatch import fnmatch
import os
def main():
filedir = './StrIntell'
for file in os.listdir(filedir):
if fnmatch(file, '*.jpg'):
fileLoc=filedir+"/"+file
img=cv2.imread(fileLoc)
# img=cv2.copyMakeBorder(img,10,10,10,10,cv2.BORDER_CONSTANT,value=[255,255,255]) # 扩大
# img = img[0:25, 0:25] # 裁剪 高*宽
print(img.shape)
cv2.imwrite(fileLoc, img)
if __name__ == '__main__':
main()
挑选好合适的库图片并将其resize
使用TestImgCut.py直接读库识别验证码
根据效果挑选合适的切割位置并保存起来
当觉得库文件与切割位置合适，将图片转为list并保存在变量
保存在变量的主要目的是可以直接读取到内存，避免频繁读硬盘造成时间浪费
binary.py转字符为变量
import cv2
import os
from fnmatch import fnmatch
import numpy as np
np.set_printoptions(threshold=np.inf) # 不省略输出
if __name__ == '__main__':
binary = ""
for fileImg in os.listdir("StrIntell/"):
if fnmatch(fileImg, '*.jpg'):
img = cv2.imread("StrIntell/"+fileImg,0)
binary = binary + "'" +fileImg.split(".")[0] + "'" + ":" + str(img.tolist()) + ","
# cv2.imwrite("test.jpg", np.array(img.tolist()))
binary = "charMap = {" + binary + "}"
with open("CharMap.py",'w+') as f:
f.write(binary)
CharMap.py字符变量
charMap = {'1':[[255, 255, 254, 254, 255, 251, 254, 255, 254, 253, 254, 255, 255], [252, 254, 251, 255, 255, 254, 255, 255, 254, 253, 254, 255, 255], [255, 249, 255, 252, 248, 255, 250, 255, 252, 252, 253, 254, 254], [253, 255, 250, 255, 249, 255, 1, 0, 251, 252, 253, 253, 254], [253, 255, 250, 253, 5, 1, 3, 0, 253, 254, 254, 254, 253], [254, 251, 255, 253, 0, 0, 5, 2, 253, 255, 255, 254, 253], [254, 254, 250, 255, 252, 254, 2, 0, 251, 253, 255, 255, 254], [254, 250, 255, 255, 255, 254, 3, 5, 250, 252, 254, 254, 254], [255, 255, 248, 255, 249, 254, 2, 0, 255, 255, 255, 253, 255], [252, 255, 251, 255, 255, 253, 1, 0, 254, 255, 253, 254, 255], [255, 251, 254, 255, 250, 254, 2, 0, 255, 255, 252, 253, 251], [253, 255, 252, 253, 248, 253, 0, 6, 255, 251, 254, 252, 251], [255, 250, 255, 249, 255, 255, 0, 2, 250, 255, 253, 255, 254], [254, 255, 253, 255, 0, 0, 2, 1, 1, 2, 254, 251, 255], [254, 254, 247, 255, 0, 3, 3, 0, 3, 3, 254, 251, 254], [252, 253, 255, 252, 255, 255, 251, 255, 254, 254, 253, 255, 252], [255, 255, 255, 249, 255, 253, 255, 252, 255, 255, 252, 251, 254]],'2':[[249, 255, 251, 254, 255, 253, 253, 253, 255, 255, 252, 251, 255], [255, 253, 255, 251, 249, 255, 254, 255, 252, 253, 255, 255, 253], [253, 254, 252, 255, 254, 253, 255, 253, 253, 255, 250, 252, 255], [254, 255, 252, 2, 0, 3, 1, 0, 255, 255, 253, 254, 255], [254, 252, 5, 0, 2, 0, 3, 1, 3, 249, 253, 255, 254], [254, 255, 254, 249, 251, 251, 253, 253, 4, 1, 254, 255, 251], [255, 254, 251, 255, 251, 255, 255, 250, 0, 4, 253, 251, 254], [255, 250, 251, 252, 255, 246, 253, 254, 9, 2, 252, 255, 251], [248, 255, 253, 252, 255, 255, 255, 5, 0, 254, 253, 254, 255], [253, 255, 251, 255, 252, 0, 0, 0, 255, 251, 255, 251, 255], [255, 250, 252, 255, 1, 2, 255, 253, 250, 255, 252, 255, 250], [254, 253, 255, 0, 6, 255, 247, 255, 252, 255, 252, 251, 255], [254, 255, 2, 2, 0, 250, 255, 253, 251, 254, 253, 252, 255], [254, 254, 3, 2, 1, 1, 5, 1, 3, 1, 255, 253, 252], [252, 251, 1, 3, 0, 3, 0, 4, 7, 1, 252, 254, 255], [254, 255, 255, 255, 255, 255, 255, 253, 252, 255, 254, 255, 253], [252, 255, 255, 255, 253, 255, 251, 253, 255, 255, 251, 255, 254]],'3':[[255, 253, 253, 255, 255, 253, 251, 255, 254, 253, 255, 255, 250], [255, 253, 251, 255, 255, 254, 253, 249, 255, 254, 253, 255, 254], [253, 251, 255, 252, 252, 255, 254, 255, 255, 254, 253, 252, 253], [255, 253, 249, 253, 255, 0, 4, 0, 0, 0, 255, 255, 252], [254, 255, 255, 250, 3, 3, 0, 6, 4, 0, 0, 255, 254], [254, 255, 255, 253, 252, 255, 252, 252, 253, 3, 0, 254, 255], [255, 250, 255, 255, 252, 253, 255, 253, 253, 254, 0, 248, 255], [255, 255, 254, 254, 253, 253, 253, 252, 255, 0, 2, 253, 253], [254, 255, 253, 255, 255, 251, 0, 1, 0, 3, 253, 255, 253], [254, 250, 255, 255, 255, 255, 0, 5, 4, 0, 0, 252, 254], [255, 255, 254, 252, 255, 254, 254, 255, 255, 0, 3, 254, 253], [254, 250, 255, 254, 254, 254, 253, 250, 251, 255, 0, 255, 255], [255, 255, 253, 255, 255, 252, 253, 255, 255, 4, 3, 251, 251], [255, 255, 252, 254, 254, 4, 0, 1, 4, 0, 2, 255, 254], [254, 254, 255, 252, 254, 0, 0, 2, 5, 0, 255, 250, 254], [255, 255, 254, 254, 255, 254, 255, 254, 255, 253, 253, 252, 251], [255, 255, 255, 255, 255, 255, 254, 254, 252, 255, 253, 255, 254]],'b':[[254, 255, 255, 255, 254, 255, 253, 255, 255, 254, 255, 255, 253], [255, 252, 251, 253, 252, 255, 254, 252, 255, 255, 255, 252, 255], [253, 255, 255, 252, 255, 255, 252, 255, 255, 250, 255, 255, 255], [255, 253, 0, 1, 252, 252, 255, 252, 253, 255, 253, 254, 255], [255, 251, 4, 0, 255, 252, 255, 254, 255, 253, 255, 253, 251], [253, 255, 0, 5, 254, 254, 255, 253, 249, 250, 255, 255, 253], [253, 255, 4, 1, 3, 0, 1, 0, 9, 254, 250, 249, 255], [254, 253, 2, 1, 0, 3, 0, 5, 0, 1, 255, 249, 253], [253, 251, 4, 0, 4, 255, 255, 252, 1, 0, 3, 255, 252], [255, 255, 0, 4, 250, 249, 255, 255, 255, 0, 0, 254, 254], [252, 254, 0, 5, 254, 255, 252, 252, 255, 5, 0, 255, 255], [254, 253, 4, 0, 255, 251, 250, 255, 254, 1, 2, 255, 254], [254, 255, 0, 0, 2, 255, 254, 252, 3, 0, 1, 253, 255], [248, 253, 1, 4, 2, 0, 2, 4, 1, 0, 255, 253, 255], [255, 255, 4, 1, 253, 2, 4, 0, 13, 249, 254, 255, 252], [249, 255, 254, 251, 255, 253, 254, 253, 254, 255, 253, 255, 251], [255, 254, 255, 251, 255, 255, 253, 252, 252, 255, 255, 255, 255]],'c':[[254, 255, 255, 255, 255, 254, 254, 255, 255, 255, 254, 255, 253], [255, 255, 251, 254, 255, 255, 255, 255, 254, 255, 254, 254, 254], [255, 255, 255, 252, 255, 251, 254, 254, 255, 253, 255, 254, 255], [254, 251, 255, 255, 254, 255, 251, 254, 253, 255, 254, 254, 255], [255, 255, 252, 254, 255, 250, 255, 253, 255, 248, 255, 255, 255], [255, 255, 255, 252, 251, 255, 255, 251, 255, 254, 255, 255, 250], [249, 255, 255, 252, 7, 0, 0, 2, 0, 255, 251, 255, 255], [255, 252, 253, 7, 0, 3, 0, 0, 0, 255, 255, 254, 254], [254, 255, 1, 5, 2, 254, 254, 254, 255, 249, 255, 255, 254], [252, 255, 0, 6, 247, 255, 252, 255, 253, 254, 254, 254, 255], [255, 250, 0, 0, 255, 255, 252, 255, 254, 255, 251, 253, 255], [254, 252, 4, 1, 252, 255, 252, 250, 251, 254, 255, 255, 255], [250, 255, 0, 4, 0, 250, 254, 255, 255, 250, 255, 254, 249], [255, 255, 254, 0, 1, 0, 2, 0, 0, 252, 254, 255, 255], [254, 255, 252, 255, 3, 0, 0, 3, 2, 255, 252, 255, 255], [248, 255, 252, 253, 254, 255, 255, 255, 253, 255, 255, 255, 250], [255, 255, 254, 251, 255, 253, 252, 254, 255, 253, 255, 255, 254]],'m':[[254, 253, 255, 252, 255, 252, 255, 255, 255, 255, 253, 255, 255], [255, 255, 252, 255, 252, 255, 253, 254, 252, 255, 255, 252, 255], [255, 255, 255, 253, 255, 254, 254, 255, 253, 255, 254, 254, 255], [254, 253, 254, 255, 255, 254, 251, 253, 255, 255, 253, 255, 253], [254, 255, 255, 251, 254, 254, 253, 253, 253, 252, 254, 253, 255], [255, 250, 255, 255, 255, 255, 252, 255, 254, 254, 255, 255, 255], [255, 255, 0, 8, 253, 0, 7, 0, 5, 251, 250, 255, 254], [254, 255, 1, 0, 2, 9, 1, 1, 1, 4, 1, 255, 255], [255, 253, 6, 0, 1, 254, 255, 255, 3, 0, 1, 255, 252], [255, 251, 1, 0, 255, 255, 249, 254, 0, 3, 255, 250, 255], [254, 253, 2, 1, 252, 254, 252, 255, 3, 0, 255, 254, 252], [255, 255, 0, 1, 255, 252, 255, 253, 0, 7, 253, 249, 255], [254, 251, 4, 0, 250, 254, 255, 254, 2, 0, 255, 255, 252], [255, 255, 2, 3, 254, 255, 254, 255, 4, 0, 255, 253, 255], [254, 255, 0, 0, 255, 253, 253, 255, 1, 0, 255, 254, 248], [255, 254, 255, 255, 253, 255, 255, 255, 253, 255, 253, 255, 255], [255, 253, 251, 252, 254, 254, 254, 255, 254, 255, 255, 254, 254]],'n':[[254, 255, 253, 252, 255, 255, 252, 255, 254, 255, 253, 255, 255], [255, 253, 255, 255, 252, 252, 255, 255, 255, 255, 255, 255, 254], [255, 254, 255, 255, 254, 255, 250, 253, 251, 255, 255, 254, 255], [255, 255, 253, 255, 253, 255, 255, 255, 255, 254, 250, 255, 252], [254, 255, 255, 252, 255, 254, 254, 253, 251, 255, 254, 255, 255], [255, 254, 255, 253, 253, 255, 254, 255, 255, 254, 254, 252, 253], [254, 254, 7, 0, 255, 254, 0, 0, 0, 5, 254, 255, 251], [253, 255, 0, 1, 1, 0, 8, 0, 4, 0, 1, 252, 254], [254, 255, 0, 4, 2, 0, 251, 255, 245, 0, 1, 250, 254], [253, 254, 0, 2, 0, 255, 254, 252, 252, 1, 0, 255, 252], [252, 251, 5, 0, 253, 254, 255, 251, 255, 2, 1, 253, 255], [255, 250, 2, 6, 250, 255, 250, 255, 250, 0, 2, 255, 249], [247, 255, 0, 0, 254, 253, 255, 254, 255, 2, 0, 255, 255], [250, 255, 3, 1, 255, 255, 252, 255, 250, 6, 1, 254, 253], [255, 252, 3, 0, 255, 254, 251, 253, 254, 0, 0, 255, 255], [253, 255, 253, 255, 253, 255, 255, 255, 253, 255, 255, 251, 253], [255, 253, 251, 251, 254, 251, 255, 254, 254, 255, 252, 253, 255]],'v':[[255, 255, 254, 255, 253, 255, 252, 255, 255, 254, 255, 255, 253], [255, 255, 254, 255, 253, 251, 255, 255, 254, 255, 254, 252, 255], [255, 254, 255, 255, 255, 254, 255, 254, 253, 253, 255, 255, 254], [253, 255, 254, 252, 254, 255, 251, 255, 254, 255, 254, 254, 254], [255, 253, 255, 254, 255, 255, 254, 254, 255, 255, 254, 255, 255], [255, 255, 254, 248, 254, 250, 254, 255, 255, 250, 255, 252, 252], [252, 255, 252, 255, 254, 255, 252, 253, 255, 255, 1, 253, 253], [254, 255, 253, 253, 6, 0, 254, 250, 255, 0, 1, 255, 253], [253, 252, 255, 253, 1, 0, 255, 254, 251, 3, 0, 255, 252], [253, 252, 251, 255, 0, 3, 254, 251, 255, 3, 1, 252, 255], [255, 255, 254, 253, 255, 0, 0, 255, 0, 1, 255, 254, 251], [255, 251, 252, 251, 248, 1, 0, 4, 1, 2, 254, 254, 255], [255, 252, 254, 255, 255, 0, 3, 0, 3, 0, 255, 249, 253], [254, 254, 253, 255, 254, 254, 4, 0, 0, 255, 251, 255, 255], [255, 255, 255, 255, 255, 252, 2, 0, 1, 248, 255, 254, 248], [255, 253, 255, 254, 255, 255, 255, 252, 255, 255, 252, 252, 254], [255, 255, 254, 255, 255, 253, 255, 254, 254, 251, 255, 255, 252]],'x':[[255, 255, 255, 255, 253, 255, 253, 255, 255, 255, 255, 255, 254], [253, 255, 254, 255, 254, 255, 255, 255, 252, 255, 255, 251, 255], [255, 253, 255, 252, 255, 255, 250, 255, 255, 255, 255, 255, 252], [255, 254, 254, 253, 253, 255, 254, 255, 253, 253, 254, 254, 255], [254, 255, 253, 255, 254, 250, 252, 255, 255, 255, 253, 255, 255], [250, 255, 253, 252, 254, 255, 254, 254, 252, 255, 255, 251, 255], [255, 252, 0, 4, 254, 251, 255, 4, 2, 255, 250, 255, 253], [251, 252, 5, 0, 255, 254, 254, 0, 3, 255, 255, 251, 254], [255, 249, 255, 9, 0, 251, 14, 0, 255, 255, 254, 255, 255], [251, 255, 252, 253, 0, 4, 0, 255, 254, 253, 251, 251, 255], [252, 255, 254, 254, 5, 3, 5, 255, 248, 255, 255, 255, 255], [255, 252, 254, 252, 1, 4, 0, 255, 253, 255, 249, 251, 255], [254, 255, 252, 6, 1, 250, 1, 1, 255, 251, 255, 255, 253], [252, 255, 1, 0, 255, 255, 255, 3, 5, 251, 255, 252, 255], [255, 252, 4, 2, 254, 251, 253, 2, 0, 254, 255, 253, 253], [255, 251, 254, 255, 254, 255, 252, 255, 255, 255, 252, 254, 254], [255, 253, 252, 252, 253, 255, 253, 251, 255, 253, 254, 255, 251]],'z':[[255, 255, 255, 254, 255, 254, 255, 255, 255, 255, 255, 254, 255], [254, 254, 255, 255, 255, 254, 255, 253, 255, 255, 255, 254, 254], [255, 255, 252, 253, 252, 255, 255, 255, 255, 252, 255, 255, 255], [255, 255, 252, 255, 254, 248, 255, 250, 254, 255, 249, 255, 254], [255, 253, 255, 254, 255, 255, 255, 253, 253, 254, 254, 254, 254], [253, 253, 255, 252, 250, 250, 251, 253, 255, 254, 251, 255, 255], [255, 254, 0, 3, 6, 4, 9, 0, 0, 255, 252, 251, 254], [253, 253, 2, 0, 0, 3, 0, 1, 1, 250, 255, 253, 254], [253, 255, 254, 252, 255, 4, 0, 1, 255, 254, 251, 255, 255], [255, 255, 254, 253, 255, 2, 0, 254, 254, 252, 255, 253, 255], [253, 247, 255, 252, 4, 6, 252, 255, 255, 254, 255, 253, 252], [255, 255, 252, 9, 0, 254, 250, 250, 252, 254, 255, 255, 253], [255, 254, 0, 0, 5, 254, 255, 255, 255, 254, 253, 254, 251], [255, 252, 3, 4, 3, 0, 0, 1, 0, 254, 254, 254, 255], [248, 255, 3, 0, 2, 1, 1, 0, 1, 255, 252, 254, 255], [255, 250, 255, 254, 254, 255, 255, 255, 254, 253, 254, 255, 249], [252, 253, 255, 253, 254, 255, 252, 253, 255, 255, 255, 255, 255]],}
Convert.py转换为灰度图并降噪
import cv2
import numpy as np
class Convert(object):
"""docstring for Convert"""
def __init__(self):
super(Convert, self).__init__()
def _get_dynamic_binary_image(self,img):
'''
自适应阀值二值化
'''
img = cv2.imdecode(np.frombuffer(img, np.uint8), cv2.IMREAD_COLOR)
img = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
th1 = cv2.adaptiveThreshold(img, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY, 21, 1)
return th1
def clear_border(self,img):
'''去除边框
'''
h, w = img.shape[:2]
for y in range(0, w):
for x in range(0, h):
# if y ==0 or y == w -1 or y == w - 2:
if y < 4 or y > w -4:
img[x, y] = 255
# if x == 0 or x == h - 1 or x == h - 2:
if x < 4 or x > h - 4:
img[x, y] = 255
return img
def interference_line(self,img):
'''
干扰线降噪
'''
h, w = img.shape[:2]
# ！！！opencv矩阵点是反的
# img[1,2] 1:图片的高度，2：图片的宽度
for y in range(1, w - 1):
for x in range(1, h - 1):
count = 0
if img[x, y - 1] > 245:
count = count + 1
if img[x, y + 1] > 245:
count = count + 1
if img[x - 1, y] > 245:
count = count + 1
if img[x + 1, y] > 245:
count = count + 1
if count > 2:
img[x, y] = 255
return img
def interference_point(self,img, x = 0, y = 0):
"""点降噪
9邻域框,以当前点为中心的田字框,黑点个数
:param x:
:param y:
:return:
"""
# todo 判断图片的长宽度下限
cur_pixel = img[x,y]# 当前像素点的值
height,width = img.shape[:2]
for y in range(0, width - 1):
for x in range(0, height - 1):
if y == 0: # 第一行
if x == 0: # 左上顶点,4邻域
# 中心点旁边3个点
sum = int(cur_pixel) \
+ int(img[x, y + 1]) \
+ int(img[x + 1, y]) \
+ int(img[x + 1, y + 1])
if sum <= 2 * 245:
img[x, y] = 0
elif x == height - 1: # 右上顶点
sum = int(cur_pixel) \
+ int(img[x, y + 1]) \
+ int(img[x - 1, y]) \
+ int(img[x - 1, y + 1])
if sum <= 2 * 245:
img[x, y] = 0
else: # 最上非顶点,6邻域
sum = int(img[x - 1, y]) \
+ int(img[x - 1, y + 1]) \
+ int(cur_pixel) \
+ int(img[x, y + 1]) \
+ int(img[x + 1, y]) \
+ int(img[x + 1, y + 1])
if sum <= 3 * 245:
img[x, y] = 0
elif y == width - 1: # 最下面一行
if x == 0: # 左下顶点
# 中心点旁边3个点
sum = int(cur_pixel) \
+ int(img[x + 1, y]) \
+ int(img[x + 1, y - 1]) \
+ int(img[x, y - 1])
if sum <= 2 * 245:
img[x, y] = 0
elif x == height - 1: # 右下顶点
sum = int(cur_pixel) \
+ int(img[x, y - 1]) \
+ int(img[x - 1, y]) \
+ int(img[x - 1, y - 1])
if sum <= 2 * 245:
img[x, y] = 0
else: # 最下非顶点,6邻域
sum = int(cur_pixel) \
+ int(img[x - 1, y]) \
+ int(img[x + 1, y]) \
+ int(img[x, y - 1]) \
+ int(img[x - 1, y - 1]) \
+ int(img[x + 1, y - 1])
if sum <= 3 * 245:
img[x, y] = 0
else: # y不在边界
if x == 0: # 左边非顶点
sum = int(img[x, y - 1]) \
+ int(cur_pixel) \
+ int(img[x, y + 1]) \
+ int(img[x + 1, y - 1]) \
+ int(img[x + 1, y]) \
+ int(img[x + 1, y + 1])
if sum <= 3 * 245:
img[x, y] = 0
elif x == height - 1: # 右边非顶点
sum = int(img[x, y - 1]) \
+ int(cur_pixel) \
+ int(img[x, y + 1]) \
+ int(img[x - 1, y - 1]) \
+ int(img[x - 1, y]) \
+ int(img[x - 1, y + 1])
if sum <= 3 * 245:
img[x, y] = 0
else: # 具备9领域条件的
sum = int(img[x - 1, y - 1]) \
+ int(img[x - 1, y]) \
+ int(img[x - 1, y + 1]) \
+ int(img[x, y - 1]) \
+ int(cur_pixel) \
+ int(img[x, y + 1]) \
+ int(img[x + 1, y - 1]) \
+ int(img[x + 1, y]) \
+ int(img[x + 1, y + 1])
if sum <= 4 * 245:
img[x, y] = 0
return img
def run(self,img):
# 自适应阈值二值化
img = self._get_dynamic_binary_image(img)
# 去除边框
img = self.clear_border(img)
# 对图片进行干扰线降噪
img = self.interference_line(img)
# 对图片进行点降噪
img = self.interference_point(img)
return img
ImgMain.py识别代码
#!/usr/bin/python
# -*- coding: utf-8 -*-
from fnmatch import fnmatch
from queue import Queue
import matplotlib.pyplot as plt
import cv2
import time
import os
from Convert import Convert
from CharMap import charMap
import requests
import numpy as np
def cutting_img(im,im_position,xoffset = 1,yoffset = 1):
# 识别出的字符个数
im_number = len(im_position[1])
if(im_number>=4): im_number = 4;
imgArr = []
# 切割字符
for i in range(im_number):
im_start_X = im_position[1][i][0] - xoffset
im_end_X = im_position[1][i][1] + xoffset
im_start_Y = im_position[2][i][0] - yoffset
im_end_Y = im_position[2][i][1] + yoffset
cropped = im[im_start_Y:im_end_Y, im_start_X:im_end_X]
imgArr.append(cropped)
# cv2.imwrite(str(i)+"v.jpg",cropped) # 查看切割效果
return im_number,imgArr
def main():
cvt = Convert()
req = requests.get("http://xxxxxxxxxxxxxxxx/verifycode.servlet")
# 注意有些教务加装了所谓云防护，没有请求头会拦截，导致获取不了验证码图片，报错可以打印req.content看看
img = cvt.run(req.content)
cv2.imwrite("v.jpg",img) # 查看验证码
#切割的位置
im_position = ([8, 7, 6, 9], [[4, 12], [14, 21], [24, 30], [34, 43]], [[7, 16], [7, 16], [7, 16], [7, 16]])
cutting_img_num,imgArr = cutting_img(img,im_position,1,1)
# 识别验证码
result=""
for i in range(cutting_img_num):
try:
template = imgArr[i]
tempResult=""
matchingDegree=0.0
for char in charMap:
img = np.asarray(charMap[char],dtype = np.uint8)
res = cv2.matchTemplate(img,template,3) #img原图 template模板 用模板匹配原图
min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(res)
if(max_val>matchingDegree):
tempResult=char
matchingDegree=max_val
result += tempResult
matchingDegree=0.0
except Exception as err:
raise Exception
# print("ERROR "+ str(err))
pass
print(result)
if __name__ == '__main__':
main()
提供全部代码
https://github.com/WindrunnerMax/SWVerifyCode
强智教务系统验证码识别 OpenCV的更多相关文章
- 强智教务系统验证码识别 Tensorflow CNN
强智教务系统验证码识别 Tensorflow CNN 一直都是使用API取得数据,但是API提供的数据较少,且为了防止API关闭,先把验证码问题解决 使用Tensorflow训练模型,强智教务系统的验 ...
- Java模拟登录带验证码的教务系统（原理详解）
一:原理 客户端访问服务器,服务器通过Session对象记录会话,服务器可以指定一个唯一的session ID作为cookie来代表每个客户端,用来识别这个客户端接下来的请求. 我们通过Chrome浏 ...
- uu云验证码识别平台,验证码,验证码识别,全自动验证码识别技术,优优云全自动打码,代答题系统,优优云远程打码平台,uu云打码
uu云验证码识别平台,验证码,验证码识别,全自动验证码识别技术,优优云全自动打码,代答题系统,优优云远程打码平台,uu云打码 优优云验证码识别答题平台介绍 优优云|UU云(中国公司)是全球唯一领先的智 ...
- 使用tensorflow搭建自己的验证码识别系统
目录 准备验证码数据 保存为tfrecords文件 验证码训练 学习tensorflow有一段时间了,想做点东西来练一下手.为了更有意思点,下面将搭建一个简单的验证码识别系统. 准备验证码数据 下面将 ...
- windows下简单验证码识别——完美验证码识别系统
此文已由作者徐迪授权网易云社区发布. 欢迎访问网易云社区,了解更多网易技术产品运营经验. 讲到验证码识别,大家第一个可能想到tesseract.诚然,对于OCR而言,tesseract确实很强大,自带 ...
- 字符型图片验证码识别完整过程及Python实现
字符型图片验证码识别完整过程及Python实现 1 摘要 验证码是目前互联网上非常常见也是非常重要的一个事物,充当着很多系统的 防火墙 功能,但是随时OCR技术的发展,验证码暴露出来的安全问题也越 ...
- 完整的验证码识别流程基于svm（若是想提升，可优化）
字符型图片验证码识别完整过程及Python实现 首先很感觉这篇文章的作者,将这篇文章写的这么好.我呢,也是拿来学习,觉得太好,所以忍不住就进行了转载. 因为我个人现在手上也有个验证码识别的项目,只是难 ...
- 字符识别Python实现 图片验证码识别
字符型图片验证码识别完整过程及Python实现 1 摘要 验证码是目前互联网上非常常见也是非常重要的一个事物,充当着很多系统的 防火墙 功能,但是随时OCR技术的发展,验证码暴露出来的安全问题也越 ...
- 基于tensorflow的‘端到端’的字符型验证码识别源码整理(github源码分享)
基于tensorflow的‘端到端’的字符型验证码识别 1 Abstract 验证码(CAPTCHA)的诞生本身是为了自动区分 自然人 和 机器人 的一套公开方法, 但是近几年的人工智能技术的发展 ...
随机推荐
- Dangling meta character '' near index 0
1.replaceAll()方法报错 对字符串使用replaceAll()方法替换 * ? + / | 等字符的时候会报以下异常 Dangling meta character '*' near in ...
- python调用adb命令进行手机操作
Python中执行cmd命令可以用到os和subprocess两个模块. 区别在于os是阻塞式的,subprocess是非阻塞式的,所以一般我们使用subprocess是比较适合的. 接下来我先举一个 ...
- Dart-Tour2-类
类 Dart语法样式: https://www.dartlang.org/guides/language/effective-dart/style语法:https://www.dartlang.org ...
- 14、创建/恢复ETH钱包身份
借助网上的一段描述: 若以银行账户为类比,这 5 个词分别对应内容如下: 地址=银行卡号密码=银行卡密码私钥=银行卡号+银行卡密码助记词=银行卡号+银行卡密码Keystore+密码=银行卡号+银行卡密 ...
- 用新架构适配MI3中遇到的各种坑
用新架构适配MI3中遇到的各种坑 首先不得不说hendy架构的强大之处, mi3也直接开机但是遇到各种坑,不能怪架构不够强大,只有说miui定制化太高.下面详细说一下mi3适配中的各种坑.有些坑会附带 ...
- Java对接微信登录
今天我们来对接微信开放平台的网站应用登录 首先上文档链接:https://developers.weixin.qq.com/doc/oplatform/Website_App/WeChat_Login ...
- Electron打包H5网页为桌面运行程序
一.安装配置环境 Electron(一种桌面应用程序运行时),Electron 把 Chromium 和 Node 合并到一个单独的运行时里面,很适合开发桌面 web 形式的应用程序,通过Node它提 ...
- Java基础--数组的定义
1.数组的定义 数组:一组能够储存相同数据类型值的变量的集合. 2.数组的赋值方式 (1)使用默认的初始值来初始化数组中的每一个元素 语法:数组元素类型[]数组名 = new数组元素类型[数组中元素的 ...
- JZOJ 1349. 最大公约数 (Standard IO)
1349. 最大公约数 (Standard IO) Time Limits: 1000 ms Memory Limits: 65536 KB Description 小菜的妹妹小诗就要读小学了!正所谓 ...
- vue中犯下的小错误（一）
在开发采筑平台SRM的移动项目中:一个页面,感觉没啥错误,但是页面报错如下: 页面中的data或者mothods都没有任何问题,但是这个报错很是让人纠结,后来发现,在使用子组件时候: 此tabShow ...