文件掃描OCR識別-1（python）

工具包匯入

import numpy as np

import cv2

1

2

函式設定

# 四邊形座標求解

def order_points(pts):

# 一共 4 個座標點

rect = np.zeros((4, 2), dtype = "float32")

# 按順序找到對應座標 0123 分別是 左上，右上，右下，左下

# 計算左上，右下

s = pts.sum(axis = 1)

rect[0] = pts[np.argmin(s)]

rect[2] = pts[np.argmax(s)]

# 計算右上和左下

diff = np.diff(pts, axis = 1)

rect[1] = pts[np.argmin(diff)]

rect[3] = pts[np.argmax(diff)]

return rect

# 獲取輸入座標點

def four_point_transform(image, pts):

rect = order_points(pts)

(tl, tr, br, bl) = rect

# 計算輸入的 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))

# 變換後對應座標位置

dst = np.array([

[0, 0],

[maxWidth - 1, 0],

[maxWidth - 1, maxHeight - 1],

[0, maxHeight - 1]], dtype="float32")

# 計算變換矩陣  透視變換 -- 二維升三維再降維  齊次座標 : 用 N+1 維來代表 N 維座標 [kx,ky,k]

M = cv2.getPerspectiveTransform(rect, dst)

warped = cv2.warpPerspective(image, M, (maxWidth, maxHeight))

# 返回變換後結果

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

讀取輸入

image = cv2.imread('images/receipt.jpg')

1

邊緣檢測

ratio = image.shape[0] / 500.0

# image.shape[0], 圖片垂直尺寸

# image.shape[1], 圖片水平尺寸

# image.shape[2], 圖片通道數

orig = image.copy()

image = resize(orig, height=500)  # 等比例縮放

# 預處理

gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

gray = cv2.GaussianBlur(gray, (5, 5), 0)  # 去除噪音點

edged = cv2.Canny(gray, 75, 200)  # 邊緣檢測

# 展示預處理結果

print("STEP 1: 邊緣檢測 ")

cv2.imshow("Image", image)

cv2.imshow("Edged", edged)

cv2.waitKey(0)

cv2.destroyAllWindows()

獲取輪廓

# 輪廓檢測

cnts = cv2.findContours(edged.copy(), cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)[1]

cnts = sorted(cnts, key=cv2.contourArea, reverse=True)[:5]  # 選取前五個 , 最大的輪廓

# 遍歷輪廓

for c in cnts:

# 計算輪廓近似

peri = cv2.arcLength(c, True)  # retval=cv.arcLength(curve, closed) retval 返回值，輪廓的周長 closed 曲線是是否閉合

# C 表示輸入的點集

# epsilon 表示從原始輪廓到近似輪廓的最大距離，外匯跟單gendan5.com它是一個準確度引數

# True 表示封閉的

approx = cv2.approxPolyDP(c, 0.02 * peri, True)  # 輪廓 , 輪廓精度 , 越小可能是多邊形 , 越大可能是矩形

# 4 個點的時候就拿出來

if len(approx) == 4:

screenCnt = approx

# print(screenCnt)  # 四個點的座標

break

# 展示結果

print("STEP 2: 獲取輪廓 ")

cv2.drawContours(image, [screenCnt], -1, (0, 255, 0), 2)

cv2.imshow("Outline", image)

cv2.waitKey(0)

cv2.destroyAllWindows()

變換

# 透視變換

warped = four_point_transform(orig, screenCnt.reshape(4, 2) * ratio)

# 二值處理

warped = cv2.cvtColor(warped, cv2.COLOR_BGR2GRAY)

ref = cv2.threshold(warped, 100, 255, cv2.THRESH_BINARY)[1]

cv2.imwrite('scan.jpg', ref)

# 展示結果

print("STEP 3: 變換 ")

cv2.imshow("Original", resize(orig, height = 650))

cv2.imshow("Scanned", resize(ref, height = 650))

cv2.waitKey(0)