opencv-python提取二维码

主要算法：
利用二维码的三个定位点来找到提取二维码
算法的主要思想和C++参考该博主：C++ Opencv提取二维码
该博主的算法是不完整的。
我的算法思想：
1）定位点的轮廓有三层轮廓
2）每个定位点的轮廓中心点一样的
3）三个定位点可以围成一个等腰直角三角形
算法待优化的地方
1）只能识别拍摄的比较清楚的二维码，拍摄位置的角度偏差比较小
2）只能识别出一个二维码

# -*- coding=utf-8 -*-
import os
import cv2
import numpy as np
import copy
img_path='images'
img_result='results'
def reshape_image(image):
    '''归一化图片尺寸：短边400，长边不超过800，短边400，长边超过800以长边800为主'''
    width,height=image.shape[1],image.shape[0]
    min_len=width
    scale=width*1.0/400
    new_width=400
    new_height=int(height/scale)
    if new_height>800:
        new_height=800
        scale=height*1.0/800
        new_width=int(width/scale)
    out=cv2.resize(image,(new_width,new_height))
    return out
def detecte(image):
    '''提取所有轮廓'''
    gray=cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    _,gray=cv2.threshold(gray,0,255,cv2.THRESH_OTSU+cv2.THRESH_BINARY_INV)
    img,contours,hierachy=cv2.findContours(gray,cv2.RETR_TREE,cv2.CHAIN_APPROX_SIMPLE)
    return image,contours,hierachy

def compute_1(contours,i,j):
    '''最外面的轮廓和子轮廓的比例'''
    area1 = cv2.contourArea(contours[i])
    area2 = cv2.contourArea(contours[j])
    if area2==0:
        return False
    ratio = area1 * 1.0 / area2
    if abs(ratio - 49.0 / 25):
        return True
    return False
def compute_2(contours,i,j):
    '''子轮廓和子子轮廓的比例'''
    area1 = cv2.contourArea(contours[i])
    area2 = cv2.contourArea(contours[j])
    if area2==0:
        return False
    ratio = area1 * 1.0 / area2
    if abs(ratio - 25.0 / 9):
        return True
    return False
def compute_center(contours,i):
    '''计算轮廓中心点'''
    M=cv2.moments(contours[i])
    cx = int(M['m10'] / M['m00'])
    cy = int(M['m01'] / M['m00'])
    return cx,cy
def detect_contours(vec):
    '''判断这个轮廓和它的子轮廓以及子子轮廓的中心的间距是否足够小'''
    distance_1=np.sqrt((vec[0]-vec[2])**2+(vec[1]-vec[3])**2)
    distance_2=np.sqrt((vec[0]-vec[4])**2+(vec[1]-vec[5])**2)
    distance_3=np.sqrt((vec[2]-vec[4])**2+(vec[3]-vec[5])**2)
    if sum((distance_1,distance_2,distance_3))/3<3:
        return True
    return False
def juge_angle(rec):
    '''判断寻找是否有三个点可以围成等腰直角三角形'''
    if len(rec)<3:
        return -1,-1,-1
    for i in range(len(rec)):
        for j in range(i+1,len(rec)):
            for k in range(j+1,len(rec)):
                distance_1 = np.sqrt((rec[i][0] - rec[j][0]) ** 2 + (rec[i][1] - rec[j][1]) ** 2)
                distance_2 = np.sqrt((rec[i][0] - rec[k][0]) ** 2 + (rec[i][1] - rec[k][1]) ** 2)
                distance_3 = np.sqrt((rec[j][0] - rec[k][0]) ** 2 + (rec[j][1] - rec[k][1]) ** 2)
                if abs(distance_1-distance_2)<5:
                    if abs(np.sqrt(np.square(distance_1)+np.square(distance_2))-distance_3)<5:
                        return i,j,k
                elif abs(distance_1-distance_3)<5:
                    if abs(np.sqrt(np.square(distance_1)+np.square(distance_3))-distance_2)<5:
                        return i,j,k
                elif abs(distance_2-distance_3)<5:
                    if abs(np.sqrt(np.square(distance_2)+np.square(distance_3))-distance_1)<5:
                        return i,j,k
    return -1,-1,-1
def find(image,image_name,contours,hierachy,root=0):
    '''找到符合要求的轮廓'''
    rec=[]
    for i in range(len(hierachy)):
        child = hierachy[i][2]
        child_child=hierachy[child][2]
        if child!=-1 and hierachy[child][2]!=-1:
            if compute_1(contours, i, child) and compute_2(contours,child,child_child):
                cx1,cy1=compute_center(contours,i)
                cx2,cy2=compute_center(contours,child)
                cx3,cy3=compute_center(contours,child_child)
                if detect_contours([cx1,cy1,cx2,cy2,cx3,cy3]):
                    rec.append([cx1,cy1,cx2,cy2,cx3,cy3,i,child,child_child])
    '''计算得到所有在比例上符合要求的轮廓中心点'''
    i,j,k=juge_angle(rec)
    if i==-1 or j== -1 or k==-1:
        return
    ts = np.concatenate((contours[rec[i][6]], contours[rec[j][6]], contours[rec[k][6]]))
    rect = cv2.minAreaRect(ts)
    box = cv2.boxPoints(rect)
    box = np.int0(box)
    result=copy.deepcopy(image)
    cv2.drawContours(result, [box], 0, (0, 0, 255), 2)
    cv2.drawContours(image,contours,rec[i][6],(255,0,0),2)
    cv2.drawContours(image,contours,rec[j][6],(255,0,0),2)
    cv2.drawContours(image,contours,rec[k][6],(255,0,0),2)
    cv2.imshow('img',image)
    cv2.waitKey(0)
    cv2.imshow('img',result)
    cv2.waitKey(0)
    path=os.path.join(img_result,image_name)
    cv2.imwrite(path,result)
    return
if __name__ == '__main__':
    files=os.listdir(img_path)
    for file in files:
        image=cv2.imread(os.path.join(img_path,file))
        image=reshape_image(image)
        cv2.imshow('img', image)
        cv2.waitKey(0)
        image,contours,hierachy=detecte(image)
        find(image,file,contours,np.squeeze(hierachy))

结果：