![\"\"](\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/01\/20260125002431-697562bf779da.png\")
![\"\"](\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/01\/20260125002431-697562bf9188c.png\")
<\/p>\n\u4e00\u3001\u6307\u7eb9\u8bc6\u522b<\/p>\n
1.\u4e0a\u7bc7\u535a\u5ba2\u4e2d\u8bb2\u5230\u6307\u7eb9\u8bc6\u522b\u4ee5\u53ca\u5bf9\u5339\u914d\u4e0a\u7684\u70b9\u6807\u51fa，\u8fd9\u6b21\u6211\u4eec\u5b9e\u73b0\u8ba9\u5339\u914d\u5bf9\u5e94\u7684\u70b9\u8fde\u63a5\u8d77\u6765，\u4ece\u8fde\u63a5\u7684\u7ebf\u4e0a\u53cd\u5e94\u5339\u914d\u7684\u7a0b\u5ea6<\/p>\n
\u8fd9\u91cc\u548c\u4e0a\u7bc7\u535a\u5ba2\u4e2d\u4ee3\u7801\u4e00\u6837，\u53ea\u4e0d\u8fc7\u5728\u51fd\u6570\u4e2d\u7a0d\u6539\u52a8\u4e86，\u6dfb\u52a0\u4e86src_matches=[]\u8fd9\u4e2a\u53d8\u91cf，\u4ee5\u53ca\u540e\u9762\u5bf9\u5176\u7684\u4f7f\u7528<\/p>\n
import cv2
\ndef cv_show(name,img):
\n cv2.imshow(name,img)
\n cv2.waitKey(0)
\ndef verification(src,model,show_matchs=False):
\n sift=cv2.SIFT_create()
\n kp1,des1=sift.detectAndCompute(src,None)
\n kp2,des2=sift.detectAndCompute(model,None)
\n flann=cv2.FlannBasedMatcher()
\n matches=flann.knnMatch(des1,des2,k=2)
\n ok=[]
\n good_matches=[]
\n src_matches=[]
\n model_matches=[]
\n for m,n in matches:
\n if m.distance<0.4*n.distance:
\n ok.append((m,n))
\n good_matches.append(m)
\n src_matches.append(kp1[m.queryIdx].pt)
\n model_matches.append(kp2[m.trainIdx].pt)
\n num=len(ok)
\n if num >=50:
\n result='\u8ba4\u8bc1\u901a\u8fc7'
\n else:
\n result='\u8ba4\u8bc1\u5931\u8d25'
\n if show_matchs and len(src_matches)>0:
\n src_with_matches=src.copy()
\n model_with_matches=model.copy()
\n for (x,y) in src_matches[:100]:
\n cv2.circle(src_with_matches,(int(x),int(y)),3,(0,0,255),-1)
\n for (x,y) in model_matches[:100]:
\n cv2.circle(model_with_matches,(int(x),int(y)),3,(0,0,255),-1)
\n cv2.imshow('src_matches',src_with_matches)
\n cv2.imshow('model_matches',model_with_matches)
\n cv2.waitKey(0)
\n match_img=cv2.drawMatches(src,kp1,model,kp2,good_matches[:100],None,flags=cv2.DrawMatchesFlags_NOT_DRAW_SINGLE_POINTS)
\n cv2.imshow('pipei',match_img)
\n cv2.waitKey(0)#1
\n return result,num,src_matches,model_matches<\/p>\n
if __name__=='__main__':
\n src1=cv2.imread('zhiwen1.bmp')
\n src2=cv2.imread('zhiwen2.bmp')
\n model = cv2.imread('zhiwen_model.bmp')
\n cv2.imshow('src1',src1)
\n cv2.imshow('src2',src2)
\n cv2.imshow('model',model)
\n cv2.waitKey(0)
\n reselt1,num1,src1_matches,model1_matches=verification(src1,model,show_matchs=True)
\n reselt2,num2,src2_matches,model2_matches=verification(src2,model,show_matchs=True)
\n print('src1\u7684\u7ed3\u679c\u4e3a：',reselt1,'\u5339\u914d\u70b9\u6570：{}'.format(num1))
\n print('src2\u7684\u7ed3\u679c\u4e3a：',reselt2,'\u5339\u914d\u70b9\u6570：{}'.format(num2))<\/p>\n
2.\u6307\u7eb9\u8bc6\u522b\u5e76\u8f93\u51fa\u5bf9\u5e94\u4eba\u540d<\/p>\n
\u8fd9\u6bb5\u548c\u4e4b\u524d\u6307\u7eb9\u8bc6\u522b\u7684\u662f\u4e00\u6837\u7684<\/p>\n
import os
\nimport cv2<\/p>\n
# \u8ba1\u7b97\u4e24\u4e2a\u6307\u7eb9\u95f4\u5339\u914d\u70b9\u7684\u4e2a\u6570
\ndef getNum(src, model):
\n img1 = cv2.imread(src)
\n img2 = cv2.imread(model)
\n sift = cv2.SIFT_create() # orb_create()
\n kp1, des1 = sift.detectAndCompute(img1, None)
\n kp2, des2 = sift.detectAndCompute(img2, None)
\n flann = cv2.FlannBasedMatcher()
\n matches = flann.knnMatch(des1, des2, k=2)
\n ok = []
\n for m, n in matches:
\n if m.distance < 0.8 * n.distance:
\n ok.append(m)
\n num = len(ok)
\n return num<\/p>\n
\u5982\u679c\u8981\u5bf9\u5e94\u4e0a\u4eba\u7684\u59d3\u540d\u8fd9\u91cc\u6211\u4eec\u65b0\u589e\u4e24\u4e2a\u51fd\u6570，\u4e00\u662f\u5728\u8fd9\u8fd9\u4e2a\u51fd\u6570\u91cc\u9762\u8fdb\u884c\u6307\u7eb9\u5339\u914d，\u83b7\u53d6\u5339\u914d\u6210\u529f\u8fd9\u4e2a\u6307\u7eb9\u56fe\u7247\u7684\u7f16\u53f7，\u4e8c\u662f\u521b\u5efa\u5b57\u5178\u6bcf\u4e2a\u7f16\u53f7\u5bf9\u5e94\u4e00\u4e2a\u4eba<\/p>\n
# \u83b7\u53d6\u6307\u7eb9\u7f16\u53f7
\ndef getID(src, database):
\n max = 0
\n for file in os.listdir(database):
\n model = os.path.join(database, file)
\n num = getNum(src, model)
\n print("\u6587\u4ef6\u540d:", file, "\u5339\u914d\u70b9\u4e2a\u6570:", num)
\n if num > max:
\n max = num
\n name = file
\n ID = name[0]
\n if max < 200: # src\u56fe\u7247\u4e0d\u4e00\u5b9a\u662f\u5e93\u91cc\u9762\u4eba\u7684\u6307\u7eb9
\n ID = 9999
\n return ID<\/p>\n
# \u6839\u636e\u6307\u7eb9\u7f16\u53f7，\u83b7\u53d6\u5bf9\u5e94\u59d3\u540d
\ndef getName(ID):
\n nameID = {0: '\u5f20\u4e09', 1: '\u674e\u56db', 2: '\u738b\u4e94', 3: '\u8d75\u516d', 4: '\u6731\u8001\u4e03', 5: '\u94b1\u516b', 6: '\u66f9\u4e5d', 7: '\u738b\u4e8c\u9ebb\u5b50', 8: 'andy', 9: 'Anna', 9999: '\u6ca1\u627e\u5230'}
\n name = nameID.get(int(ID))
\n return name<\/p>\n
\u6307\u7eb9\u8bc6\u522b，\u548c\u54ea\u4e2a\u7f16\u53f7\u7684\u6307\u7eb9\u5bf9\u5e94\u4e0a\u5c31\u8c03\u7528\u51fd\u6570\u5bf9\u5e94\u7684\u4e2a\u4eba\u540d\u5b57\u3002<\/p>\n
# \u4e3b\u51fd\u6570
\nif __name__ == "__main__":
\n src = "src.bmp"
\n database = "database"
\n ID = getID(src, database)
\n name = getName(ID)
\n print("\u8bc6\u522b\u7ed3\u679c\u4e3a:", name)<\/p>\n
\u4e8c\u3001\u56fe\u7247\u62fc\u63a5<\/p>\n
\u00a0 \u00a0 \u00a0 \u00a0 \u00a0<\/p>\n
\u76ee\u7684\u8ba9\u8fd9\u4e24\u5f20\u56fe\u65e0\u7f1d\u8854\u63a5\u6210\u4e00\u5f20\u56fe\u7247<\/p>\n
\u505a\u51c6\u5907：<\/p>\n
import cv2
\nimport numpy as np
\nimport sys<\/p>\n
def cv_show(name, img):
\n cv2.imshow(name, img)
\n cv2.waitKey(0)<\/p>\n
def detectAndDescribe(image):
\n gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY) # \u5c06\u5f69\u8272\u56fe\u7247\u8f6c\u6362\u6210\u7070\u5ea6\u56fe
\n sift = cv2.SIFT_create() # \u5efa\u7acbSIFT\u751f\u6210\u5668
\n # \u68c0\u6d4bSIFT\u7279\u5f81\u70b9，\u5e76\u8ba1\u7b97\u63cf\u8ff0\u7b26，\u7b2c\u4e8c\u4e2a\u53c2\u6570\u4e3a\u63a9\u819c
\n (kps, des) = sift.detectAndCompute(gray, None)
\n # \u5c06\u7ed3\u679c\u8f6c\u6362\u6210NumPy\u6570\u7ec4
\n kps_float = np.float32([kpc.pt for kpc in kps])
\n # kpc \u5305\u542b\u4e24\u4e2a\u503c，\u5206\u522b\u662f\u5173\u952e\u70b9\u5728\u56fe\u50cf\u4e2d\u7684 x \u548c y \u5750\u6807\u3002\u8fd9\u4e9b\u5750\u6807\u901a\u5e38\u662f\u6d6e\u70b9\u6570，\u53ef\u4ee5\u7cbe\u786e\u5730\u63cf\u8ff0\u5173\u952e\u70b9\u5728\u56fe\u50cf\u4e2d\u7684\u4f4d\u7f6e\u3002
\n return (kps, kps_float, des) # \u8fd4\u56de\u7279\u5f81\u70b9\u96c6，\u53ca\u5bf9\u5e94\u7684\u63cf\u8ff0\u7279\u5f81<\/p>\n
'''\u8bfb\u53d6\u62fc\u63a5\u56fe\u7247'''
\nimageA = cv2.imread("zuo.jpg")
\ncv_show('imageA', imageA)
\nimageB = cv2.imread("you.jpg")
\ncv_show('imageB', imageB)<\/p>\n
'''\u8ba1\u7b97\u56fe\u7247\u7279\u5f81\u70b9\u53ca\u63cf\u8ff0\u7b26'''
\n(kpsA, kps_floatA, desA) = detectAndDescribe(imageA)
\n(kpsB, kps_floatB, desB) = detectAndDescribe(imageB)<\/p>\n
'''\u5efa\u7acb\u5339\u914d\u5668BFMatcher，\u5728\u5339\u914d\u5927\u56fe\u8bad\u7ec3\u96c6\u5408\u65f6\u4f7f\u7528FlannBasedMatcher\u901f\u5ea6\u66f4\u5feb\u3002'''
\nmatcher = cv2.BFMatcher()
\nrawMatches = matcher.knnMatch(desB, desA, k=2)
\ngood = []
\nmatches = []
\nfor m in rawMatches:
\n # \u5f53\u6700\u8fd1\u8ddd\u79bb\u8ddf\u6b21\u8fd1\u8ddd\u79bb\u7684\u6bd4\u503c\u5c0f\u4e8e0.65\u65f6，\u4fdd\u7559\u6b64\u5339\u914d\u5bf9
\n if len(m) == 2 and m[0].distance < 0.65 * m[1].distance:
\n good.append(m)
\n # \u5b58\u50a8\u4e24\u4e2a\u70b9\u5728featuresA\u3001featuresB\u4e2d\u7684\u7d22\u5f15\u503c
\n matches.append((m[0].queryIdx, m[0].trainIdx))
\nprint(len(good))
\nprint(matches)
\nvis = cv2.drawMatchesKnn(imageB, kpsB, imageA, kpsA, good, outImg=None, flags=cv2.DRAW_MATCHES_FLAGS_DRAW_RICH_KEYPOINTS)
\ncv_show("Keypoint Matches", vis)<\/p>\n
\u8ba1\u7b97\u900f\u89c6\u53d8\u6362\u77e9\u9635
\nfindHomography(srcPoints, dstPoints, method=None, ransacReprojThreshold=None, mask=None, maxIters=None, confidence=None)<\/p>\n
findHomography\u7684\u56db\u79cd\u65b9\u6cd5：<\/p>\n
findHomography\u8fd4\u56de\u503c\u89e3\u91ca：<\/p>\n
'''\u900f\u89c6\u53d8\u6362'''
\nif len(matches) > 4: # \u5f53\u7b5b\u9009\u540e\u7684\u5339\u914d\u5bf9\u5927\u4e8e4\u65f6，\u8ba1\u7b97\u89c6\u89d2\u53d8\u6362\u77e9\u9635\u3002
\n # \u83b7\u53d6\u5339\u914d\u5bf9\u7684\u70b9\u5750\u6807
\n ptsB = np.float32([kps_floatB[i] for (i, _) in matches]) # matches\u662f\u901a\u8fc7\u9608\u503c\u7b5b\u9009\u4e4b\u540e\u7684\u7279\u5f81\u70b9\u5bf9\u8c61
\n ptsA = np.float32([kps_floatA[i] for (_, i) in matches]) # kps_floatA\u662f\u56fe\u7247A\u4e2d\u7684\u5168\u90e8\u7279\u5f81\u70b9\u5750\u6807
\n H, mask = cv2.findHomography(ptsB, ptsA, cv2.RANSAC, 10)
\nelse:
\n print('\u56fe\u7247\u672a\u627e\u52304\u4e2a\u4ee5\u4e0a\u7684\u5339\u914d\u70b9')
\n sys.exit()<\/p>\n
result = cv2.warpPerspective(imageB, H, (imageB.shape[1] + imageA.shape[1], imageB.shape[0]))
\ncv_show('resultB', result)
\n# \u5c06\u56fe\u7247A\u4f20\u5165resultB\u540e\u663e\u793a
\nresult[0:imageA.shape[0], 0:imageA.shape[1]] = imageA
\ncv_show('result', result)
\ncv2.imwrite('pingjie.jpg', result)<\/p>\n
\u5c55\u793azuo\u548cyou，\u5e76\u5bf9\u5339\u914d\u4e0a\u7684\u7279\u5f81\u8fdb\u884c\u8fde\u7ebf，\u8fd9\u70b9\u548c\u6307\u7eb9\u8bc6\u522b\u662f\u4e00\u6837\u3002<\/p>\n
![\"\"](\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/01\/20260125002431-697562bfa65c8.png\")
<\/p>\n
\u5c55\u793ayou\u56fe\u7ecf\u8fc7\u900f\u89c6\u4e4b\u540e\u7684\u53d8\u5316\u4ee5\u53ca\u6700\u540e\u62fc\u63a5\u6210\u529f\u7684\u7ed3\u679c\u3002\u6211\u4eec\u80fd\u770b\u51fa\u8fd8\u662f\u6709\u62fc\u63a5\u75d5\u8ff9\u7684，\u5c31\u6bd4\u5982\u7eb8\u7bb1\u5904\u3002<\/p>\n
![\"\"](\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/01\/20260125002432-697562c0a3927.png\")
<\/p>\n
\u4e09\u3001\u601d\u8003<\/p>\n
\u4e0a\u9762\u6211\u4eec\u56fe\u50cf\u62fc\u63a5\u662f\u5bf9b\u56fe\u8fdb\u884c\u900f\u89c6\u8f6c\u5316，\u628a\u900f\u89c6\u7684b\u56fe\u62fc\u63a5\u5728a\u56fe\u7684\u53f3\u4fa7，\u601d\u8003\u6211\u4eec\u80fd\u4e0d\u80fd\u5bf9\u56fea\u8fdb\u884c\u900f\u89c6，\u628aa\u56fe\u62fc\u63a5\u5728b\u56fe\u7684\u5de6\u4fa7\u3002\u5f97\u5230\u4e0b\u9762\u8fd9\u79cd\u7ed3\u679c\u7684\u62fc\u63a5\u3002<\/p>\n
![\"\"](\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/01\/20260125002433-697562c1b7c79.png\")
<\/p>\n","protected":false},"excerpt":{"rendered":"
\u4e00\u3001\u6307\u7eb9\u8bc6\u522b1.\u4e0a\u7bc7\u535a\u5ba2\u4e2d\u8bb2\u5230\u6307\u7eb9\u8bc6\u522b\u4ee5\u53ca\u5bf9\u5339\u914d\u4e0a\u7684\u70b9\u6807\u51fa，\u8fd9\u6b21\u6211\u4eec\u5b9e\u73b0\u8ba9\u5339\u914d\u5bf9\u5e94\u7684\u70b9\u8fde\u63a5\u8d77\u6765，\u4ece\u8fde\u63a5\u7684\u7ebf\u4e0a\u53cd\u5e94\u5339\u914d\u7684\u7a0b\u5ea6\u8fd9\u91cc\u548c\u4e0a\u7bc7\u535a\u5ba2\u4e2d\u4ee3\u7801\u4e00\u6837，\u53ea\u4e0d\u8fc7\u5728\u51fd\u6570\u4e2d\u7a0d\u6539\u52a8\u4e86，\u6dfb\u52a0\u4e86src_matches[]\u8fd9\u4e2a\u53d8\u91cf，\u4ee5\u53ca\u540e\u9762\u5bf9\u5176\u7684\u4f7f\u7528import cv2
\ndef cv_show(name,img):cv2.imshow(name,img)cv2.waitKey(0)
\ndef verification(src,<\/p>\n","protected":false},"author":2,"featured_media":65531,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[1460,50,523],"topic":[],"class_list":["post-65536","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-server","tag-opencv","tag-50","tag-523"],"yoast_head":"\n