python opencv實現信用卡的數字識別
阿新 • • 發佈:2020-01-13
本專案利用python以及opencv實現信用卡的數字識別
前期準備
- 匯入工具包
- 定義功能函式
模板影象處理
- 讀取模板影象 cv2.imread(img)
- 灰度化處理 cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
- 二值化 cv2.threshold()
- 輪廓 - 輪廓
信用卡影象處理
- 讀取信用卡影象 cv2.imread(img)
- 灰度化處理 cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
- 禮帽處理 cv2.morphologyEx(gray,cv2.MORPH_TOPHAT,rectKernel)
- Sobel邊緣檢測 cv2.Sobel(tophat,ddepth=cv2.CV_32F,dx=1,dy=0,ksize=-1)
- 閉操作 cv2.morphologyEx(gradX,cv2.MORPH_CLOSE,rectKernel)
- 計算輪廓 cv2.findContours
- 模板檢測 cv2.matchTemplate(roi,digitROI,cv2.TM_CCOEFF)
原始資料展示
結果展示
1 前期準備
# 匯入工具包 # opencv讀取圖片的格式為b g r # matplotlib圖片的格式為 r g b import numpy as np import cv2 from imutils import contours import matplotlib.pyplot as plt %matplotlib inline
# 信用卡的位置 predict_card = "images/credit_card_01.png" # 模板的位置 template = "images/ocr_a_reference.png"
# 指定信用卡型別 FIRST_NUMBER = { "3": "American Express","4": "Visa","5": "MasterCard","6": "Discover Card" }
# 定義一些功能函式 # 對框進行排序 def sort_contours(cnts,method="left-to-right"): reverse = False i = 0 if method == "right-to-left" or method == "bottom-to-top": reverse = True if method == "top-to-bottom" or method == "bottom-to-top": i = 1 boundingBoxes = [cv2.boundingRect(c) for c in cnts] #用一個最小的矩形,把找到的形狀包起來x,y,h,w (cnts,boundingBoxes) = zip(*sorted(zip(cnts,boundingBoxes),key=lambda b: b[1][i],reverse=reverse)) return cnts,boundingBoxes # 調整圖片尺寸大小 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 # 定義cv2展示函式 def cv_show(name,img): cv2.imshow(name,img) cv2.waitKey(0) cv2.destroyAllWindows()
2 對模板影象進行預處理操作
讀取模板影象
# 讀取模板影象 img = cv2.imread(template) cv_show("img",img) plt.imshow(img)
<matplotlib.image.AxesImage at 0x2b2e04ad128>
模板影象轉灰度影象
# 轉灰度圖 ref = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY) cv_show("ref",ref) plt.imshow(ref)
<matplotlib.image.AxesImage at 0x2b2e25d9e48>
轉為二值影象
ref = cv2.threshold(ref,10,255,cv2.THRESH_BINARY_INV)[1] cv_show("ref",ref) plt.imshow(ref)
<matplotlib.image.AxesImage at 0x2b2e2832a90>
計算輪廓
#cv2.findContours()函式接受的引數為二值圖,即黑白的(不是灰度圖),cv2.RETR_EXTERNAL只檢測外輪廓,cv2.CHAIN_APPROX_SIMPLE只保留終點座標 #返回的list中每個元素都是影象中的一個輪廓 # 在二值化後的影象中計算輪廓 refCnts,hierarchy = cv2.findContours(ref.copy(),cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_SIMPLE) # 在原圖上畫出輪廓 cv2.drawContours(img,refCnts,-1,(0,255),3) cv_show("img",img) plt.imshow(img)
<matplotlib.image.AxesImage at 0x2b2e256f908>
print(np.array(refCnts).shape) # 排序,從左到右,從上到下 refCnts = sort_contours(refCnts,method="left-to-right")[0] digits = {} # 遍歷每一個輪廓 for (i,c) in enumerate(refCnts): # 計算外接矩形並且resize成合適大小 (x,w,h) = cv2.boundingRect(c) roi = ref[y:y + h,x:x + w] roi = cv2.resize(roi,(57,88)) # 每一個數字對應每一個模板 digits[i] = roi
(10,)
3 對信用卡進行處理
初始化卷積核
rectKernel = cv2.getStructuringElement(cv2.MORPH_RECT,(9,3)) sqKernel = cv2.getStructuringElement(cv2.MORPH_RECT,(5,5))
讀取信用卡
image = cv2.imread(predict_card) cv_show("image",image) plt.imshow(image)
<matplotlib.image.AxesImage at 0x2b2e294c9b0>
對影象進行預處理操作
# 先對影象進行resize操作 image = resize(image,width=300) # 灰度化處理 gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY) cv_show("gray",gray) plt.imshow(gray)
<matplotlib.image.AxesImage at 0x2b2e255d828>
對影象禮帽操作
- 禮帽 = 原始輸入-開運算結果
- 開運算:先腐蝕,再膨脹
- 突出更明亮的區域
tophat = cv2.morphologyEx(gray,rectKernel) cv_show("tophat",tophat) plt.imshow(tophat)
<matplotlib.image.AxesImage at 0x2b2eb008e48>
用Sobel運算元邊緣檢測
gradX = cv2.Sobel(tophat,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) cv_show("gradX",gradX) plt.imshow(gradX)
(189,300) <matplotlib.image.AxesImage at 0x2b2e0797400>
對影象閉操作
- 閉操作:先膨脹,再腐蝕
- 可以將數字連在一起
gradX = cv2.morphologyEx(gradX,rectKernel) cv_show("gradX",gradX) plt.imshow(gradX)
<matplotlib.image.AxesImage at 0x2b2e097cc88>
#THRESH_OTSU會自動尋找合適的閾值,適合雙峰,需把閾值引數設定為0 thresh = cv2.threshold(gradX,cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1] cv_show("thresh",thresh) plt.imshow(thresh)
<matplotlib.image.AxesImage at 0x2b2e24a0dd8>
# 再進行一次閉操作 thresh = cv2.morphologyEx(thresh,sqKernel) #再來一個閉操作 cv_show("thresh",thresh) plt.imshow(thresh)
<matplotlib.image.AxesImage at 0x2b2e25fe748>
計算輪廓
threshCnts,hierarchy = cv2.findContours(thresh.copy(),cv2.CHAIN_APPROX_SIMPLE) cnts = threshCnts cur_img = image.copy() cv2.drawContours(cur_img,cnts,3) cv_show("img",cur_img) plt.imshow(cur_img)
<matplotlib.image.AxesImage at 0x2b2eb17c780>
locs = [] # 遍歷輪廓 for (i,c) in enumerate(cnts): # 計算矩形 (x,h) = cv2.boundingRect(c) ar = w / float(h) # 選擇合適的區域,根據實際任務來,這裡的基本都是四個數字一組 if ar > 2.5 and ar < 4.0: if (w > 40 and w < 55) and (h > 10 and h < 20): #符合的留下來 locs.append((x,h)) # 將符合的輪廓從左到右排序 locs = sorted(locs,key=lambda x:x[0]) output = []
模板匹配
# 遍歷每一個輪廓中的數字 for (i,(gX,gY,gW,gH)) in enumerate(locs): # initialize the list of group digits groupOutput = [] # 根據座標提取每一個組 group = gray[gY - 5:gY + gH + 5,gX - 5:gX + gW + 5] cv_show("group",group) # 預處理 group = cv2.threshold(group,cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1] cv_show("group",group) # 計算每一組的輪廓 digitCnts,hierarchy = cv2.findContours(group.copy(),cv2.CHAIN_APPROX_SIMPLE) digitCnts = contours.sort_contours(digitCnts,method="left-to-right")[0] # 計算每一組中的每一個數值 for c in digitCnts: # 找到當前數值的輪廓,resize成合適的的大小 (x,h) = cv2.boundingRect(c) roi = group[y:y + h,x:x + w] roi = cv2.resize(roi,88)) cv_show("roi",roi) # 計算匹配得分 scores = [] # 在模板中計算每一個得分 for (digit,digitROI) in digits.items(): # 模板匹配 result = cv2.matchTemplate(roi,cv2.TM_CCOEFF) (_,score,_,_) = cv2.minMaxLoc(result) scores.append(score) # 得到最合適的數字 groupOutput.append(str(np.argmax(scores))) # 畫出來 cv2.rectangle(image,(gX - 5,gY - 5),(gX + gW + 5,gY + gH + 5),1) cv2.putText(image,"".join(groupOutput),gY - 15),cv2.FONT_HERSHEY_SIMPLEX,0.65,2) # 得到結果 output.extend(groupOutput)
# 列印結果 print("Credit Card Type: {}".format(FIRST_NUMBER[output[0]])) print("Credit Card #: {}".format("".join(output))) cv_show("Image",image) plt.imshow(image)
Credit Card Type: Visa Credit Card #: 4000123456789010 <matplotlib.image.AxesImage at 0x2b2eb040748>
以上就是本文的全部內容,希望對大家的學習有所幫助,也希望大家多多支援我們。