# 导入工具包
from imutils import contours
import imutils
import numpy as np
import argparse
import cv2
# 设置参数
ap = argparse.ArgumentParser()
ap.add_argument("-i", "--image", default='./images/1.png',
help="path to input image")
ap.add_argument("-t", "--template", default='./images/template.png',
help="path to template OCR-A image")
ap.add_argument("-s", "--save", default=True,
help="save every image")
args = vars(ap.parse_args())
# 定义了一个字典 FIRST_NUMBER ，它将第一个数字映射到相应的信用卡类型。
FIRST_NUMBER = {
"3": "American Express",
"4": "Visa",
"5": "MasterCard",
"6": "Discover Card"
}
def cv_show(name, image):
cv2.imshow(name, image)
cv2.waitKey(0)
cv2.destroyAllWindows()
def sort_contours(origin_counters, method="left-to-right"):
reverse = False
order = 0
if method == "right-to-left" or method == "bottom-to-top":
reverse = True
if method == "top-to-bottom" or method == "bottom-to-top":
order = 1
boxes = [cv2.boundingRect(counter) for counter in origin_counters] # x,y,h,w
ordered_counters, boxes = zip(*sorted(zip(origin_counters, boxes), key=lambda b: b[1][order], reverse=reverse))
return counters, boxes
def resize(origin_image, width=None, height=None, inter=cv2.INTER_AREA):
(h, w) = origin_image.shape[:2]
if width is None and height is None:
return origin_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_image = cv2.resize(origin_image, dim, interpolation=inter)
return resized_image
# -------------------------------------------读取模板图像,把所有数字模板放到一个数组
img = cv2.imread(args["template"])
cv2.imwrite('1template.jpg', img) if args["save"] else cv_show('template', img)
# 灰度图
ref = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
cv2.imwrite('2gray.jpg', ref) if args["save"] else cv_show('gray', ref)
# 二值图像
ref = cv2.threshold(ref, 10, 255, cv2.THRESH_BINARY_INV)[1]
cv2.imwrite('3binary.jpg', ref) if args["save"] else cv_show('binary', ref)
counters, hierarchy = cv2.findContours(ref.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
cv2.drawContours(img, counters, 1, (0, 0, 255), 3)
cv2.imwrite('4counters.jpg', img) if args["save"] else cv_show('counters', img)
# counters = sort_contours(counters, method="left-to-right")[0] # 排序，从左到右，从上到下
counters = contours.sort_contours(counters, method="left-to-right")[0] # 排序，从左到右，从上到下
templateDigits = []
for (i, c) in enumerate(counters):
# 计算外接矩形并且resize成合适大小
(x, y, w, h) = cv2.boundingRect(c)
roi = ref[y:y + h, x:x + w]
templateDigits.append(cv2.resize(roi, (57, 88)))
# 初始化几个结构化内核，构造了两个这样的内核 - 一个矩形和一个正方形。
# 我们将使用矩形的一个用于Top-hat形态运算符，将方形一个用于闭操作。
rectKernel = cv2.getStructuringElement(cv2.MORPH_RECT, (9, 3))
sqKernel = cv2.getStructuringElement(cv2.MORPH_RECT, (5, 5))
# 读取输入图像，预处理
image = cv2.imread(args["image"])
cv2.imwrite('5image.jpg', image) if args["save"] else cv_show('image', image)
image = resize(image, width=300)
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
cv2.imwrite('6gray2.jpg', gray) if args["save"] else cv_show('gray2', gray)
# 礼帽操作，突出更明亮的区域
tophat = cv2.morphologyEx(gray, cv2.MORPH_TOPHAT, rectKernel)
cv2.imwrite('7tophat.jpg', tophat) if args["save"] else cv_show('tophat', tophat)
# 计算沿x方向的渐变在计算gradX 数组中每个元素的绝对值之后 ，
# 我们采取一些步骤将值缩放到范围[0-255]（因为图像当前是浮点数据类型）。
# 要做到这一点，我们计算 最小/最大归一化 最后一步是将gradX转换 为 uint8 ，其范围为[0-255]
gradX = cv2.Sobel(tophat, ddepth=cv2.CV_32F, dx=1, dy=0, # ksize=-1相当于用3*3的
ksize=-1)
gradX = np.absolute(gradX)
(minVal, maxVal) = (np.min(gradX), np.max(gradX))
gradX = (255 * ((gradX - minVal) / (maxVal - minVal)))
gradX = gradX.astype("uint8")
print(np.array(gradX).shape)
cv2.imwrite('8gradX.jpg', gradX) if args["save"] else cv_show('gradX', gradX)
# 通过闭操作（先膨胀，再腐蚀）将数字连在一起
gradX = cv2.morphologyEx(gradX, cv2.MORPH_CLOSE, rectKernel)
cv2.imwrite('9close.jpg', gradX) if args["save"] else cv_show('close', gradX)
# THRESH_OTSU会自动寻找合适的阈值，适合双峰，需把阈值参数设置为0
thresh = cv2.threshold(gradX, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]
cv2.imwrite('10thresh.jpg', thresh) if args["save"] else cv_show('thresh', thresh)
# 再来一个闭操作
thresh = cv2.morphologyEx(thresh, cv2.MORPH_CLOSE, sqKernel) # 再来一个闭操作
cv2.imwrite('11close2.jpg', thresh) if args["save"] else cv_show('close2', thresh)
# 计算轮廓
threshCnts, hierarchy = cv2.findContours(thresh.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
cur_img = image.copy()
cv2.drawContours(cur_img, threshCnts, -1, (0, 0, 255), 3)
cv2.imwrite('12cur_img.jpg', cur_img) if args["save"] else cv_show('cur_img', cur_img)
locs = []
# 遍历轮廓
for (i, c) in enumerate(threshCnts):
# 计算矩形
(x, y, w, h) = cv2.boundingRect(c)
ar = w / float(h)
# 选择合适的区域，因为信用卡使用了一个固定大小的字体与4组4位数字，我们可以基于高宽比
if 2.5 < ar < 4.0:
# 轮廓可以进一步调整最小/最大宽度
if (40 < w < 55) and (10 < h < 20):
# 符合的留下来
locs.append((x, y, w, h))
# 将符合的轮廓从左到右排序
locs = sorted(locs, key=lambda x: x[0])
output = []
# 遍历每一个轮廓中的数字
for (i, (gX, gY, gW, gH)) in enumerate(locs):
groupOutput = []
# 根据坐标提取每一个组
group = gray[gY - 5:gY + gH + 5, gX - 5:gX + gW + 5]
cv2.imwrite('13-{}gray.jpg'.format(i), group) if args["save"] else cv_show('gray', group)
# 预处理
group = cv2.threshold(group, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]
cv2.imwrite('14-{}threshold.jpg'.format(i), img) if args["save"] else cv_show('threshold', group)
# 计算每一组的轮廓
digitCnts, hierarchy = cv2.findContours(group.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
digitCnts = contours.sort_contours(digitCnts, method="left-to-right")[0]
# 计算每一组中的每一个数值
for j, c in enumerate(digitCnts):
# 找到当前数值的轮廓，resize成合适的的大小
(x, y, w, h) = cv2.boundingRect(c)
roi = group[y:y + h, x:x + w]
roi = cv2.resize(roi, (57, 88))
cv2.imwrite('15-{}-{}roi.jpg'.format(i, j), roi) if args["save"] else cv_show('roi', roi)
scores = []
for digit in templateDigits:
# 模板匹配
result = cv2.matchTemplate(roi, digit, cv2.TM_CCOEFF)
(_, score, _, _) = cv2.minMaxLoc(result)
scores.append(score)
# 得到最合适的数字
groupOutput.append(str(np.argmax(scores)))
# 画出来
cv2.rectangle(image, (gX - 5, gY - 5),
(gX + gW + 5, gY + gH + 5), (0, 0, 255), 1)
cv2.putText(image, "".join(groupOutput), (gX, gY - 15),
cv2.FONT_HERSHEY_SIMPLEX, 0.65, (0, 0, 255), 2)
# 得到结果
output.extend(groupOutput)
# 打印结果
print("Credit Card #: {}".format("".join(output)))
cv2.imshow("Image", image)
cv2.waitKey(0)
