opencv学习笔记五--文件扫描+OCR文字识别
文件扫描定义函数边缘检测获取轮廓变换OCR文字识别环境配置代码文件扫描
# 导入工具包import numpy as npimport argparseimport cv2import 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 imageif 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] / 500.0print(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 = approxbreak
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-v5.0.1.2028.exe
https://digi.bib.uni-mannheim.de/tesseract/配置环境变量如E:\Program Files (x86)\Tesseract-OCRtesseract -v进行测试tesseract XXX.png 得到结果pip install pytesseractanaconda lib site-packges pytesseract pytesseract.pytesseract_cmd 修改为绝对路径即可
代码
from PIL import Imageimport pytesseractimport cv2import osimport 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' #threshimage = 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 KKWHOLE FOODS MARKET - WESTPORT, CT 06880399 POST RD WEST - (203) 227-6858365365365365BROTHBACONBACONBACONBACUNLSLSLSLSCHICFLOUR ALMUNDCHKN BRST BNLSS SKHEAVY CREAMBALSMC REDUCTGRND 85/15BEEFJUICECOF CASHEWLDOCS PINT ORGANICHNY ALMOND BUTTERxeene TAX.00BALNPNPNPNPNPNPNPNPNPNPNPNPNP44499.99.99mal7 7 Tmana Ramm
