python opencv 单条形码检测识别
基于python opencv pyzbar实现。
原图：
结果图：
直接上代码：
import cv2
import numpy as np
import math
from pyzbar.pyzbar import decode
#读图
img=cv2.imread(r'XX.png',1) #读取图片
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY) #将图片变为灰度图片
#使用Sobel算子,求得水平和垂直方向灰度图像的梯度差
gradX = cv2.Sobel(gray,ddepth = cv2.CV_32F,dx = 1,dy = 0,ksize = -1)
gradY = cv2.Sobel(gray,ddepth = cv2.CV_32F,dx = 0,dy = 1,ksize = -1)
gradient = cv2.subtract(gradX, gradY)
gradient = cv2.convertScaleAbs(gradient)
images = cv2.resize(gradient,(480,640))
cv2.imshow("closed1", images)
cv2.waitKey(0)
#均值滤波，消除高频噪声 (8*8)像素块
blurred = cv2.blur(gradient,(8,8))
#二值化
ret,thresh = cv2.threshold(blurred, 225, 255, cv2.THRESH_BINARY)
#闭运算 填充条码空隙 （参数自己调整）
kernel = cv2.getStructuringElement(cv2.MORPH_RECT,(25,25))
closed1 = cv2.morphologyEx(thresh,cv2.MORPH_CLOSE,kernel, iterations = 2)
# images = cv2.resize(closed1,(480,640))
# cv2.imshow("closed1", images)
# cv2.waitKey(0)
#4次腐蚀 再4次膨胀：消除小斑点
closed2 = cv2.erode(closed1, None, iterations = 4)
closed3 = cv2.dilate(closed2, None, iterations = 4)
# images = cv2.resize(closed3,(480,640))
# cv2.imshow("closed3", images)
# cv2.waitKey(0)
#开运算
ret,th2=cv2.threshold(closed3,0.1,255,cv2.THRESH_BINARY)
kernel = np.ones((10,10),np.uint8)
opening = cv2.morphologyEx(th2, cv2.MORPH_OPEN, kernel,iterations = 2)
images = cv2.resize(opening,(480,640))
cv2.imshow("contours", images)
cv2.waitKey(0)
#腐蚀
kernel = np.ones((5,5),np.uint8)
erosion = cv2.erode(opening,kernel,iterations = 2)
opening = cv2.resize(erosion ,(480,640))
cv2.imshow("contours",opening )
cv2.waitKey(0)
#找出边界
contours, hierarchy = cv2.findContours(erosion.copy(),cv2.RETR_TREE,cv2.CHAIN_APPROX_SIMPLE)
#获取最大轮廓
c = sorted(contours, key = cv2.contourArea, reverse = True)[0]
#图形旋转矫正 并截取最大边界区域
#修改最小框大小 放大50
rect = cv2.minAreaRect(c)
b = list(rect)
b[1] = list(b[1])
b[1][0] = b[1][0] + 50
b[1][1] = b[1][1] + 50
new_rect = tuple(b)
box = np.int0(cv2.boxPoints(new_rect))
#draw_img = cv2.drawContours(img.copy(), [box], -1, (0, 0, 255), 3)
# 获取画框宽高(x=orignal_W,y=orignal_H)
orignal_W = math.ceil(np.sqrt((box[3][1] - box[2][1])**2 + (box[3][0] - box[2][0])**2))
orignal_H= math.ceil(np.sqrt((box[3][1] - box[0][1])**2 + (box[3][0] - box[0][0])**2))
# 原图中的四个顶点,与变换矩阵
pts1 = np.float32([box[0], box[1], box[2], box[3]])
pts2 = np.float32([[int(orignal_W+1),int(orignal_H+1)], [0, int(orignal_H+1)], [0, 0], [int(orignal_W+1), 0]])
# 生成透视变换矩阵；进行透视变换
M = cv2.getPerspectiveTransform(pts1, pts2)
result_img = cv2.warpPerspective(img, M*1.5, (int(orignal_W+20),int(orignal_H+20)))
cv2.imshow("JIAOZHEN",result_img)
cv2.waitKey(0)
#二值化 筛掉一些错误
ret,thresh = cv2.threshold(result_img,128, 255, cv2.THRESH_BINARY)
#二维码编码检测
barcodes = decode(thresh)
result = []
for barcode in barcodes:
barcodeData = barcode.data.decode("utf-8")
result.append(barcodeData)
print(result)
最后会输出条码：
['6936281406719']
注：pyzbar模块可以识别水平竖直方向的条码，上述图像经过透视变换后基本都为竖直方向或者水平方向可以被识别出来，但是大角度的图像imshow效果不是很好，可能是倒着的。。。