# 信用卡類型及卡號OCR系統(tǒng)
# USAGE
# python ocr_template_match.py --reference images/ocr_a_reference.png --image images/credit_card_05.png

import argparse

import cv2
import imutils
import numpy as np
# 導(dǎo)入必要的包
from imutils import contours

# 構(gòu)建命令行參數(shù)及解析
# --image 必須 要進(jìn)行OCR的輸入圖像
# --reference 必須 參考OCR-A圖像
ap = argparse.ArgumentParser()
ap.add_argument("-i", "--image", required=True,
                help="path to input image")
ap.add_argument("-r", "--reference", required=True,
                help="path to reference OCR-A image")
args = vars(ap.parse_args())

# 定義一個(gè)字典（映射信用卡第一位數(shù)字和信用卡類型的編號）
FIRST_NUMBER = {
    "3": "American Express",
    "4": "Visa",
    "5": "MasterCard",
    "6": "Discover Card"
}

# 從磁盤加載參考OCR-A圖像，轉(zhuǎn)換為灰度圖，閾值化圖像以顯示為白色前景和黑色背景
# 并反轉(zhuǎn)圖像
# and invert it, such that the digits appear as *white* on a *black*
ref_origin = cv2.imread(args["reference"])
cv2.imshow("ref_origin", ref_origin)
ref = ref_origin.copy()
ref = cv2.cvtColor(ref, cv2.COLOR_BGR2GRAY)
cv2.imshow("ref_gray", ref)
ref = cv2.threshold(ref, 180, 255, cv2.THRESH_BINARY)[1]
cv2.imshow("ref_threshhold", ref)
cv2.waitKey(0)

# 尋找OCR-A圖像中的輪廓（數(shù)字的外輪廓線）
# 并從左到右排序輪廓，初始化一個(gè)字典來存儲(chǔ)數(shù)字ROI
refCnts = cv2.findContours(ref.copy(), cv2.RETR_EXTERNAL,
                           cv2.CHAIN_APPROX_SIMPLE)
print('findContours: ', len(refCnts))
refCnts = imutils.grab_contours(refCnts)
refCnts = contours.sort_contours(refCnts, method="left-to-right")[0]
digits = {}

# 遍歷OCR-A輪廓
for (i, c) in enumerate(refCnts):
    # 計(jì)算數(shù)字的邊界框，提取它，縮放到固定的大小
    (x, y, w, h) = cv2.boundingRect(c)
    cv2.rectangle(ref_origin, (x, y), (x + w, y + h), (0, 255, 0), 2)
    roi = ref[y:y + h, x:x + w]
    roi = cv2.resize(roi, (57, 88))

    # 更新數(shù)字字典，數(shù)字匹配ROI
    digits[i] = roi
cv2.imshow("ref and digits", ref_origin)
cv2.waitKey(0)

# 初始化矩形和方形結(jié)構(gòu)內(nèi)核
# 在圖像上滑動(dòng)它來進(jìn)行（卷積）操作，如模糊、銳化、邊緣檢測或其他圖像處理操作。
# 使用矩形函數(shù)作為Top-hat形態(tài)學(xué)運(yùn)算符，使用方形函數(shù)作為閉合運(yùn)算。
rectKernel = cv2.getStructuringElement(cv2.MORPH_RECT, (9, 3))
sqKernel = cv2.getStructuringElement(cv2.MORPH_RECT, (5, 5))

# 準(zhǔn)備進(jìn)行OCR的輸入圖像
# 加載輸入圖像，保持縱橫比縮放圖像寬度為300，轉(zhuǎn)換為灰度圖
origin = cv2.imread(args["image"])
origin = imutils.resize(origin, width=300)
image = origin.copy()
cv2.imshow("origin", origin)
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
cv2.imshow("gray", gray)

# 執(zhí)行形態(tài)學(xué)操作
# 應(yīng)用tophat（白帽）形態(tài)學(xué)操作以在暗的背景中提取出亮的區(qū)域（信用卡上的數(shù)字卡號）
# Top hat操作在深色背景（即信用卡號）下顯示淺色區(qū)域
tophat = cv2.morphologyEx(gray, cv2.MORPH_TOPHAT, rectKernel)
cv2.imshow("tophat", tophat)

# 計(jì)算Scharr梯度，計(jì)算梯度值
# 在白色禮帽上，計(jì)算x方向的Scharr梯度，然后縮放到范圍[0, 255]
gradX = cv2.Sobel(tophat, ddepth=cv2.CV_32F, dx=1, dy=0, ksize=-1)
gradX = np.absolute(gradX)
(minVal, maxVal) = (np.min(gradX), np.max(gradX))
# 最小/最大歸一化, 由float轉(zhuǎn)換gradX到uint8范圍[0-255]
gradX = (255 * ((gradX - minVal) / (maxVal - minVal)))
gradX = gradX.astype("uint8")
cv2.imshow("gradient", gradX)

# 使用矩形框應(yīng)用閉合操作以幫助閉合信用卡數(shù)字之間的小的縫隙
# 應(yīng)用Otsu's閾值方法二值化圖像
gradX = cv2.morphologyEx(gradX, cv2.MORPH_CLOSE, rectKernel)
cv2.imshow("morphologyEx", gradX)
thresh = cv2.threshold(gradX, 0, 255,
                       cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]
cv2.imshow("thresh1", thresh)

# 在二值化圖像上，應(yīng)用二次閉合操作
# 再一次方形框形態(tài)學(xué)操作，幫助閉合信用卡數(shù)字區(qū)域之間的縫隙
thresh = cv2.morphologyEx(thresh, cv2.MORPH_CLOSE, sqKernel)
cv2.imshow("thresh2", thresh)

# 閾值圖像中查找輪廓，然后初始化數(shù)字位置列表
cnts = cv2.findContours(thresh.copy(), cv2.RETR_EXTERNAL,
                        cv2.CHAIN_APPROX_SIMPLE)
cnts = imutils.grab_contours(cnts)
locs = []

# 遍歷輪廓
for (i, c) in enumerate(cnts):
    # 計(jì)算輪廓的邊界框，并計(jì)算縱橫比
    (x, y, w, h) = cv2.boundingRect(c)
    ar = w / float(h)

    # 由于信用卡有固定的4組4數(shù)字，可以根據(jù)縱橫比來尋找潛在的輪廓
    if ar > 2.5 and ar < 4.0:
        # 輪廓可以在最小/最大寬度上進(jìn)一步修剪
        if (w > 40 and w < 55) and (h > 10 and h < 20):
            # 添加數(shù)字組輪廓的編輯框輪廓到位置list
            locs.append((x, y, w, h))
            cv2.rectangle(origin, (x, y), (x + w, y + h), (255, 0, 0), -1)

cv2.imshow("contours filter", origin)
# 突出顯示信用卡上四組四位數(shù)字（總共十六位）。
# 從左到右排序輪廓，并初始化list來存儲(chǔ)信用卡數(shù)字列表
locs = sorted(locs, key=lambda x: x[0])
output = []

# 遍歷四組四位數(shù)字
for (i, (gX, gY, gW, gH)) in enumerate(locs):
    # 初始化存放每組數(shù)字的list
    groupOutput = []

    # 提取每組4位數(shù)字的灰度圖ROI
    # 應(yīng)用閾值方法從背景信用卡中分割數(shù)字
    group = gray[gY - 5:gY + gH + 5, gX - 5:gX + gW + 5]
    group = cv2.threshold(group, 0, 255,
                          cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]

    # 檢測組中每個(gè)單獨(dú)數(shù)字的輪廓
    # 從左到右排序輪廓
    digitCnts = cv2.findContours(group.copy(), cv2.RETR_EXTERNAL,
                                 cv2.CHAIN_APPROX_SIMPLE)
    digitCnts = imutils.grab_contours(digitCnts)
    digitCnts = contours.sort_contours(digitCnts,
                                       method="left-to-right")[0]

    # 遍歷數(shù)字輪廓
    for c in digitCnts:
        # 計(jì)算每個(gè)單獨(dú)數(shù)字的邊界框
        # 提取數(shù)字，縮放以擁有和參考OCR-A字體模板圖像相同的大小
        (x, y, w, h) = cv2.boundingRect(c)
        roi = group[y:y + h, x:x + w]
        roi = cv2.resize(roi, (57, 88))

        # 初始化模板匹配分?jǐn)?shù)list
        scores = []

        # 遍歷參考數(shù)字名和數(shù)字ROI
        for (digit, digitROI) in digits.items():
            # 應(yīng)用基于相關(guān)性的模板匹配，計(jì)算分?jǐn)?shù)，更新分?jǐn)?shù)list
            # apply correlation-based template matching, take the
            # score, and update the scores list
            result = cv2.matchTemplate(roi, digitROI,
                                       cv2.TM_CCOEFF)
            (_, score, _, _) = cv2.minMaxLoc(result)
            scores.append(score)

        # 數(shù)字ROI的分類將取 模板匹配分?jǐn)?shù)中分?jǐn)?shù)最大的參考數(shù)字
        # the classification for the digit ROI will be the reference
        # digit name with the *largest* template matching score
        groupOutput.append(str(np.argmax(scores)))

    # 圍繞每組畫一個(gè)矩形，并以紅色文本標(biāo)識(shí)圖像上的信用卡號
    # 繪制每組的數(shù)字識(shí)別分類結(jié)果
    cv2.rectangle(image, (gX - 5, gY - 5),
                  (gX + gW + 5, gY + gH + 5), (0, 0, 255), 2)
    cv2.putText(image, "".join(groupOutput), (gX, gY - 15),
                cv2.FONT_HERSHEY_SIMPLEX, 0.65, (0, 0, 255), 2)

    # 更新輸出數(shù)字分組列表
    # Pythonic的方法是使用extend函數(shù)，它將iterable對象的每個(gè)元素（本例中為列表）追加到列表的末尾
    output.extend(groupOutput)

# 顯示檢測到的信用卡類型和卡號到屏幕上
print("Credit Card Type: {}".format(FIRST_NUMBER[output[0]]))
print("Credit Card #: {}".format("".join(output)))
cv2.imshow("Image", image)
cv2.waitKey(0)