opencv学习笔记五--文件扫描+OCR文字识别

文件扫描

# 导入工具包
import numpy as np
import argparse
import cv2
import matplotlib.pyplot as plt#Matplotlib是RGB

定义函数

# 绘图展示
def cv_show(name,img):
    b,g,r = cv2.split(img)
    img_rgb = cv2.merge((r,g,b))
    plt.imshow(img_rgb)
    plt.show()
def cv_show1(name,img):
    plt.imshow(img)
    plt.show()
    cv2.imshow(name,img)
    cv2.waitKey()
    cv2.destroyAllWindows()

def order_points(pts):
    	# 一共4个坐标点
	rect = np.zeros((4, 2), dtype = "float32")

	# 按顺序找到对应坐标0123分别是 左上，右上，右下，左下
	# 计算左上，右下
	s = pts.sum(axis = 1)#横纵坐标相加，最大的是右下，最小的是左上
	rect[0] = pts[np.argmin(s)]
	rect[2] = pts[np.argmax(s)]

	# 计算右上和左下
	diff = np.diff(pts, axis = 1)
	rect[1] = pts[np.argmin(diff)]
	rect[3] = pts[np.argmax(diff)]

	return rect

def four_point_transform(image, pts):
    	# 获取输入坐标点
	rect = order_points(pts)#pts是原图上的四个点坐标
	(tl, tr, br, bl) = rect

	# 计算输入的w和h值
	widthA = np.sqrt(((br[0] - bl[0]) ** 2) + ((br[1] - bl[1]) ** 2))
	widthB = np.sqrt(((tr[0] - tl[0]) ** 2) + ((tr[1] - tl[1]) ** 2))
	maxWidth = max(int(widthA), int(widthB))

	heightA = np.sqrt(((tr[0] - br[0]) ** 2) + ((tr[1] - br[1]) ** 2))
	heightB = np.sqrt(((tl[0] - bl[0]) ** 2) + ((tl[1] - bl[1]) ** 2))
	maxHeight = max(int(heightA), int(heightB))

	# 变换后对应坐标位置
	dst = np.array([
		[0, 0],
		[maxWidth - 1, 0],
		[maxWidth - 1, maxHeight - 1],
		[0, maxHeight - 1]], dtype = "float32")

	# 计算变换矩阵
	M = cv2.getPerspectiveTransform(rect, dst)#从rect到dst的变换矩阵
	warped = cv2.warpPerspective(image, M, (maxWidth, maxHeight))#得到变换结果

	# 返回变换后结果
	return warped

def resize(image, width=None, height=None, inter=cv2.INTER_AREA):
	dim = None
	(h, w) = image.shape[:2]
	if width is None and height is None:
		return image
	if width is None:
		r = height / float(h)
		dim = (int(w * r), height)
	else:
		r = width / float(w)
		dim = (width, int(h * r))
	resized = cv2.resize(image, dim, interpolation=inter)
	return resized

# 读取输入
image = cv2.imread("./images/receipt.jpg")
#坐标也会相同变化
cv_show("Image",image)
ratio = image.shape[0] / 500.0
print(image.shape[0])
orig = image.copy()

2448

image = resize(orig, height = 500)
cv_show("Image",image)

边缘检测

# 预处理
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
gray = cv2.GaussianBlur(gray, (5, 5), 0)
edged = cv2.Canny(gray, 75, 200)

# 展示预处理结果
print("STEP 1: 边缘检测")
cv_show("Image", image)
cv_show1("Edged", edged)

STEP 1: 边缘检测

# 轮廓检测
cnts = cv2.findContours(edged.copy(), cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)[1]#检测轮廓
cnts = sorted(cnts, key = cv2.contourArea, reverse = True)[:5]#对轮廓按照面积大小进行排序
#cv2.drawContours(image, cnts, -1, (0, 255, 0), 2)
#cv_show("Outline", image)

# 遍历轮廓
for c in cnts:
	# 计算轮廓近似
	peri = cv2.arcLength(c, True)#计算轮廓周长
	# C表示输入的点集
	# epsilon表示从原始轮廓到近似轮廓的最大距离，它是一个准确度参数
	# True表示封闭的
	approx = cv2.approxPolyDP(c, 0.02 * peri, True)

	# 4个点的时候就拿出来
	if len(approx) == 4:
		screenCnt = approx
		break

获取轮廓

# 展示结果
print("STEP 2: 获取轮廓")
cv2.drawContours(image, [screenCnt], -1, (0, 255, 0), 2)
cv_show("Outline", image)

STEP 2: 获取轮廓

# 透视变换
warped = four_point_transform(orig, screenCnt.reshape(4, 2) * ratio)
print(screenCnt.reshape(4, 2))
print(screenCnt.reshape(4, 2).sum(axis = 1))

[[465 110]
 [113 137]
 [147 375]
 [474 323]]
[575 250 522 797]

变换

# 二值处理
warped = cv2.cvtColor(warped, cv2.COLOR_BGR2GRAY)
ref = cv2.threshold(warped, 100, 255, cv2.THRESH_BINARY)[1]
cv2.imwrite('scan.jpg', ref)
# 展示结果
print("STEP 3: 变换")
cv_show("Original", resize(orig, height = 650))
cv_show1("Scanned", resize(ref, height = 650))

STEP 3: 变换

OCR文字识别

环境配置

安装tesseract-ocr-w64-setup-v5.0.1.20220118.exe

1. https://digi.bib.uni-mannheim.de/tesseract/
2. 配置环境变量如E:\Program Files (x86)\Tesseract-OCR
3. tesseract -v进行测试
4. tesseract XXX.png 得到结果
5. pip install pytesseract
6. anaconda lib site-packges pytesseract pytesseract.py
7. tesseract_cmd 修改为绝对路径即可

代码

from PIL import Image
import pytesseract
import cv2
import os
import matplotlib.pyplot as plt#Matplotlib是RGB

# 绘图展示
def cv_show(name,img):
    b,g,r = cv2.split(img)
    img_rgb = cv2.merge((r,g,b))
    plt.imshow(img_rgb)
    plt.show()
def cv_show1(name,img):
    plt.imshow(img)
    plt.show()
    cv2.imshow(name,img)
    cv2.waitKey()
    cv2.destroyAllWindows()

preprocess = 'thresh' #thresh

image = cv2.imread('scan.jpg')
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

if preprocess == "thresh":
    gray = cv2.threshold(gray, 0, 255,cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]#自适应二值化

if preprocess == "blur":
    gray = cv2.medianBlur(gray, 3)#中位模糊

filename = "{}.png".format(os.getpid())
cv2.imwrite(filename, gray)

True

text = pytesseract.image_to_string(Image.open(filename))
print(text)
os.remove(filename)

cv_show("Image", image)
cv_show1("Output", gray)

we KK Re KK KK OK OK KK

WHOLE FOODS MARKET - WESTPORT, CT 06880
399 POST RD WEST - [phone]

365
365
365
365

BROTH

BACON
BACON
BACON
BACUN

LS
LS
LS
LS

CHIC

FLOUR ALMUND
CHKN BRST BNLSS SK
HEAVY CREAM
BALSMC REDUCT
GRND 85/15

BEEF

JUICE

COF CASHEW

L

DOCS PINT ORGANIC
HNY ALMOND BUTTER

xeene TAX

.00

BAL

NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP

4
4
4

99

.99
.99

mal

7 7 T

mana Ramm