{"id":1003108,"date":"2024-12-27T10:14:29","date_gmt":"2024-12-27T02:14:29","guid":{"rendered":"https:\/\/docs.pingcode.com\/ask\/ask-ask\/1003108.html"},"modified":"2024-12-27T10:14:31","modified_gmt":"2024-12-27T02:14:31","slug":"python-opencv-%e5%a6%82%e4%bd%95%e5%8e%bb%e6%b0%b4%e5%8d%b0","status":"publish","type":"post","link":"https:\/\/docs.pingcode.com\/ask\/1003108.html","title":{"rendered":"python opencv \u5982\u4f55\u53bb\u6c34\u5370"},"content":{"rendered":"

\"python<\/p>\n

Python OpenCV\u53bb\u9664\u6c34\u5370\u7684\u65b9\u6cd5\u6709\u591a\u79cd\uff0c\u5e38\u89c1\u7684\u5305\u62ec\u4f7f\u7528inpAI<\/a>nting\u6280\u672f\u3001\u56fe\u50cf\u4fee\u8865\u7b97\u6cd5\u3001\u8499\u7248\u5904\u7406<\/strong>\u7b49\u3002\u9996\u5148\uff0cinpainting\u6280\u672f\u80fd\u591f\u6709\u6548\u5730\u4fee\u590d\u56fe\u50cf\u4e2d\u7684\u5c0f\u533a\u57df\uff0c\u4f8b\u5982\u6c34\u5370\u3002\u5b83\u901a\u8fc7\u5229\u7528\u5468\u56f4\u7684\u50cf\u7d20\u4fe1\u606f\u6765\u586b\u5145\u76ee\u6807\u533a\u57df\uff0c\u4ece\u800c\u8fbe\u5230\u53bb\u9664\u6c34\u5370\u7684\u6548\u679c\u3002\u5176\u6b21\uff0c\u56fe\u50cf\u4fee\u8865\u7b97\u6cd5\uff0c\u5982Telea\u7b97\u6cd5\u6216Navier-Stokes\u7b97\u6cd5\uff0c\u53ef\u4ee5\u6839\u636e\u90bb\u57df\u4fe1\u606f\u91cd\u6784\u88ab\u6c34\u5370\u8986\u76d6\u7684\u90e8\u5206\u3002\u6700\u540e\uff0c\u8499\u7248\u5904\u7406\u662f\u901a\u8fc7\u624b\u52a8\u7ed8\u5236\u8499\u7248\u6765\u6307\u5b9a\u8981\u53bb\u9664\u7684\u533a\u57df\uff0c\u7136\u540e\u8fdb\u884c\u4fee\u590d\u3002\u4e0b\u9762\u5c06\u8be6\u7ec6\u8ba8\u8bba\u5176\u4e2d\u4e00\u79cd\u65b9\u6cd5\uff0c\u5373\u4f7f\u7528inpainting\u6280\u672f\u53bb\u9664\u6c34\u5370\u3002<\/p>\n<\/p>\n

\u4e00\u3001INPAINTING\u6280\u672f<\/h3>\n<\/p>\n

Inpainting\u6280\u672f\u662f\u4e00\u79cd\u56fe\u50cf\u4fee\u590d\u6280\u672f\uff0c\u4e3b\u8981\u7528\u4e8e\u4fee\u590d\u56fe\u50cf\u4e2d\u7684\u7f3a\u635f\u533a\u57df\u3002OpenCV\u63d0\u4f9b\u4e86\u4e24\u79cdinpainting\u7b97\u6cd5\uff1aTelea\u7b97\u6cd5\u548cNavier-Stokes\u7b97\u6cd5\u3002<\/p>\n<\/p>\n

1.1 Telea\u7b97\u6cd5<\/h4>\n<\/p>\n

Telea\u7b97\u6cd5\u662f\u4e00\u79cd\u5feb\u901f\u7684inpainting\u7b97\u6cd5\uff0c\u9002\u7528\u4e8e\u5c0f\u9762\u79ef\u7684\u4fee\u590d\u3002\u5b83\u901a\u8fc7\u6269\u6563\u90bb\u8fd1\u7684\u50cf\u7d20\u503c\u6765\u586b\u5145\u76ee\u6807\u533a\u57df\u3002\u4f7f\u7528Python\u7684OpenCV\u5e93\uff0c\u53ef\u4ee5\u8f7b\u677e\u5b9e\u73b0\u8fd9\u4e00\u7b97\u6cd5\u3002<\/p>\n<\/p>\n

import cv2<\/p>\n
import numpy as np<\/p>\n
\u8bfb\u53d6\u56fe\u50cf<\/strong><\/h2>\n
img = cv2.imread('image_with_watermark.jpg')<\/p>\n
\u521b\u5efa\u63a9\u6a21\uff0c\u6807\u8bb0\u6c34\u5370\u533a\u57df<\/strong><\/h2>\n
mask = np.zeros(img.shape[:2], np.uint8)<\/p>\n
mask[100:200, 100:300] = 255  # \u8fd9\u53ea\u662f\u4e00\u4e2a\u793a\u4f8b\uff0c\u9700\u8981\u6839\u636e\u5b9e\u9645\u6c34\u5370\u4f4d\u7f6e\u8c03\u6574<\/p>\n
\u4f7f\u7528Telea\u7b97\u6cd5\u8fdb\u884cinpainting<\/strong><\/h2>\n
dst = cv2.inpaint(img, mask, 3, cv2.INPAINT_TELEA)<\/p>\n
\u4fdd\u5b58\u548c\u663e\u793a\u7ed3\u679c<\/strong><\/h2>\n
cv2.imwrite('result_telea.jpg', dst)<\/p>\n
cv2.imshow('Inpainted Image', dst)<\/p>\n
cv2.waitKey(0)<\/p>\n
cv2.destroyAllWindows()<\/p>\n
<\/code><\/pre>\n<\/p>\n
1.2 Navier-Stokes\u7b97\u6cd5<\/h4>\n<\/p>\n
Navier-Stokes\u7b97\u6cd5\u662f\u4e00\u79cd\u57fa\u4e8e\u6d41\u4f53\u52a8\u529b\u5b66\u7684inpainting\u7b97\u6cd5\uff0c\u9002\u7528\u4e8e\u590d\u6742\u7684\u4fee\u590d\u573a\u666f\u3002\u76f8\u6bd4Telea\u7b97\u6cd5\uff0c\u5b83\u80fd\u591f\u66f4\u597d\u5730\u5904\u7406\u5927\u9762\u79ef\u7684\u4fee\u590d\u3002<\/p>\n<\/p>\n
# \u4f7f\u7528Navier-Stokes\u7b97\u6cd5\u8fdb\u884cinpainting<\/p>\n
dst_ns = cv2.inpaint(img, mask, 3, cv2.INPAINT_NS)<\/p>\n
\u4fdd\u5b58\u548c\u663e\u793a\u7ed3\u679c<\/strong><\/h2>\n
cv2.imwrite('result_ns.jpg', dst_ns)<\/p>\n
cv2.imshow('Inpainted Image NS', dst_ns)<\/p>\n
cv2.waitKey(0)<\/p>\n
cv2.destroyAllWindows()<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e8c\u3001\u56fe\u50cf\u4fee\u8865\u7b97\u6cd5<\/h3>\n<\/p>\n
\u56fe\u50cf\u4fee\u8865\u7b97\u6cd5\u662f\u53e6\u4e00\u79cd\u53bb\u9664\u6c34\u5370\u7684\u6280\u672f\uff0c\u5b83\u901a\u8fc7\u5206\u6790\u56fe\u50cf\u7684\u7eb9\u7406\u548c\u7ed3\u6784\u4fe1\u606f\uff0c\u751f\u6210\u4e0e\u539f\u59cb\u56fe\u50cf\u4e00\u81f4\u7684\u4fee\u8865\u5185\u5bb9\u3002<\/p>\n<\/p>\n
2.1 \u4f7f\u7528\u7eb9\u7406\u5408\u6210<\/h4>\n<\/p>\n
\u7eb9\u7406\u5408\u6210\u662f\u4e00\u79cd\u57fa\u4e8e\u6837\u672c\u7684\u56fe\u50cf\u4fee\u590d\u65b9\u6cd5\uff0c\u5b83\u53ef\u4ee5\u5229\u7528\u56fe\u50cf\u4e2d\u73b0\u6709\u7684\u7eb9\u7406\u4fe1\u606f\u6765\u586b\u5145\u7f3a\u5931\u90e8\u5206\u3002\u8fd9\u79cd\u65b9\u6cd5\u7279\u522b\u9002\u7528\u4e8e\u80cc\u666f\u5177\u6709\u91cd\u590d\u7eb9\u7406\u7684\u56fe\u50cf\u3002<\/p>\n<\/p>\n
2.2 PatchMatch\u7b97\u6cd5<\/h4>\n<\/p>\n
PatchMatch\u662f\u4e00\u79cd\u5feb\u901f\u7684\u56fe\u50cf\u4fee\u8865\u7b97\u6cd5\uff0c\u5b83\u901a\u8fc7\u5bfb\u627e\u6700\u4f73\u5339\u914d\u5757\u6765\u4fee\u590d\u56fe\u50cf\u3002OpenCV\u7684contrib\u6a21\u5757\u4e2d\u63d0\u4f9b\u4e86\u8fd9\u4e00\u7b97\u6cd5\u7684\u5b9e\u73b0\u3002<\/p>\n<\/p>\n
# \u4f7f\u7528PatchMatch\u7b97\u6cd5\u8fdb\u884c\u4fee\u8865\uff08\u5047\u8bbe\u5df2\u7ecf\u5b89\u88c5\u4e86\u76f8\u5e94\u7684\u6a21\u5757\uff09<\/p>\n
\u7531\u4e8ePatchMatch\u5c5e\u4e8eOpenCV\u7684\u6269\u5c55\u529f\u80fd\uff0c\u53ef\u80fd\u9700\u8981\u989d\u5916\u7684\u5e93\u652f\u6301<\/strong><\/h2>\n
<\/code><\/pre>\n<\/p>\n
\u4e09\u3001\u8499\u7248\u5904\u7406<\/h3>\n<\/p>\n
\u8499\u7248\u5904\u7406\u662f\u4e00\u79cd\u624b\u52a8\u65b9\u6cd5\uff0c\u901a\u8fc7\u7ed8\u5236\u8499\u7248\u6765\u6307\u5b9a\u8981\u53bb\u9664\u7684\u533a\u57df\uff0c\u7136\u540e\u5229\u7528\u56fe\u50cf\u4fee\u590d\u6280\u672f\u8fdb\u884c\u5904\u7406\u3002<\/p>\n<\/p>\n
3.1 \u624b\u52a8\u7ed8\u5236\u8499\u7248<\/h4>\n<\/p>\n
\u4f7f\u7528\u56fe\u50cf\u7f16\u8f91\u5de5\u5177\u624b\u52a8\u7ed8\u5236\u8499\u7248\uff0c\u6307\u5b9a\u8981\u53bb\u9664\u7684\u6c34\u5370\u533a\u57df\u3002\u8fd9\u79cd\u65b9\u6cd5\u9700\u8981\u4e00\u5b9a\u7684\u4eba\u5de5\u5e72\u9884\uff0c\u4f46\u53ef\u4ee5\u66f4\u7cbe\u786e\u5730\u63a7\u5236\u4fee\u590d\u533a\u57df\u3002<\/p>\n<\/p>\n
3.2 \u7ed3\u5408inpainting\u548c\u8499\u7248<\/h4>\n<\/p>\n
\u5c06\u624b\u52a8\u7ed8\u5236\u7684\u8499\u7248\u4e0einpainting\u6280\u672f\u7ed3\u5408\uff0c\u53ef\u4ee5\u5b9e\u73b0\u66f4\u52a0\u7cbe\u7ec6\u7684\u4fee\u590d\u6548\u679c\u3002<\/p>\n<\/p>\n
# \u7ed3\u5408\u624b\u52a8\u8499\u7248\u548cinpainting\u6280\u672f<\/p>\n
\u5047\u8bbe\u5df2\u7ecf\u521b\u5efa\u4e86\u4e00\u4e2a\u7cbe\u786e\u7684\u8499\u7248'mask_manual'<\/strong><\/h2>\n
\u4f7f\u7528Telea\u7b97\u6cd5\u8fdb\u884c\u4fee\u590d<\/strong><\/h2>\n
result = cv2.inpaint(img, mask_manual, 3, cv2.INPAINT_TELEA)<\/p>\n
\u663e\u793a\u7ed3\u679c<\/strong><\/h2>\n
cv2.imshow('Final Result', result)<\/p>\n
cv2.waitKey(0)<\/p>\n
cv2.destroyAllWindows()<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u56db\u3001\u6ce8\u610f\u4e8b\u9879<\/h3>\n<\/p>\n
\u5728\u5b9e\u9645\u64cd\u4f5c\u4e2d\uff0c\u9700\u8981\u6ce8\u610f\u4ee5\u4e0b\u51e0\u70b9\uff1a<\/p>\n<\/p>\n
4.1 \u786e\u5b9a\u6c34\u5370\u533a\u57df<\/h4>\n<\/p>\n
\u51c6\u786e\u5730\u786e\u5b9a\u6c34\u5370\u533a\u57df\u662f\u53bb\u9664\u6c34\u5370\u7684\u5173\u952e\u3002\u53ef\u4ee5\u901a\u8fc7\u624b\u52a8\u7ed8\u5236\u6216\u8005\u7f16\u7a0b\u8bc6\u522b\u6c34\u5370\u533a\u57df\uff0c\u786e\u4fdd\u4fee\u590d\u7684\u51c6\u786e\u6027\u3002<\/p>\n<\/p>\n
4.2 \u9009\u62e9\u5408\u9002\u7684\u7b97\u6cd5<\/h4>\n<\/p>\n
\u6839\u636e\u56fe\u50cf\u7684\u590d\u6742\u7a0b\u5ea6\u548c\u6c34\u5370\u7684\u6027\u8d28\uff0c\u9009\u62e9\u5408\u9002\u7684inpainting\u7b97\u6cd5\u3002\u5bf9\u4e8e\u7b80\u5355\u573a\u666f\uff0cTelea\u7b97\u6cd5\u53ef\u80fd\u8db3\u591f\uff1b\u800c\u5bf9\u4e8e\u590d\u6742\u573a\u666f\uff0cNavier-Stokes\u7b97\u6cd5\u6216PatchMatch\u53ef\u80fd\u66f4\u5408\u9002\u3002<\/p>\n<\/p>\n
4.3 \u8c03\u6574\u53c2\u6570<\/h4>\n<\/p>\n
\u5728\u4f7f\u7528\u4fee\u590d\u7b97\u6cd5\u65f6\uff0c\u9002\u5f53\u5730\u8c03\u6574\u53c2\u6570\uff08\u5982\u4fee\u590d\u534a\u5f84\uff09\u80fd\u591f\u5f71\u54cd\u4fee\u590d\u6548\u679c\u3002\u901a\u8fc7\u5b9e\u9a8c\u627e\u5230\u6700\u4f73\u7684\u53c2\u6570\u7ec4\u5408\u3002<\/p>\n<\/p>\n
\u4e94\u3001\u603b\u7ed3<\/h3>\n<\/p>\n
\u53bb\u9664\u6c34\u5370\u662f\u4e00\u4e2a\u5177\u6709\u6311\u6218\u6027\u7684\u4efb\u52a1\uff0c\u9700\u8981\u7ed3\u5408\u591a\u79cd\u6280\u672f\u624b\u6bb5\u3002\u4f7f\u7528Python\u7684OpenCV\u5e93\uff0c\u53ef\u4ee5\u901a\u8fc7inpainting\u6280\u672f\u3001\u56fe\u50cf\u4fee\u8865\u7b97\u6cd5\u548c\u8499\u7248\u5904\u7406\u7b49\u65b9\u6cd5\u5b9e\u73b0\u6c34\u5370\u7684\u53bb\u9664\u3002\u5728\u64cd\u4f5c\u8fc7\u7a0b\u4e2d\uff0c\u9700\u6ce8\u610f\u9009\u62e9\u5408\u9002\u7684\u7b97\u6cd5\u548c\u53c2\u6570\uff0c\u4ee5\u83b7\u5f97\u6700\u4f73\u7684\u4fee\u590d\u6548\u679c\u3002\u901a\u8fc7\u4e0d\u65ad\u7684\u5b9e\u8df5\u548c\u8c03\u6574\uff0c\u53ef\u4ee5\u5728\u4e0d\u540c\u7684\u573a\u666f\u4e0b\u6709\u6548\u5730\u53bb\u9664\u6c34\u5370\uff0c\u5e76\u4fdd\u6301\u56fe\u50cf\u7684\u81ea\u7136\u6027\u3002<\/p>\n<\/p>\n
\u76f8\u5173\u95ee\u7b54FAQs\uff1a<\/strong><\/h2>\n
 \u5982\u4f55\u4f7f\u7528Python\u548cOpenCV\u53bb\u9664\u6c34\u5370\u7684\u57fa\u672c\u6b65\u9aa4\u662f\u4ec0\u4e48\uff1f<\/strong>
\u53bb\u9664\u6c34\u5370\u901a\u5e38\u9700\u8981\u501f\u52a9\u56fe\u50cf\u5904\u7406\u6280\u672f\u3002\u9996\u5148\uff0c\u4f7f\u7528OpenCV\u8bfb\u53d6\u56fe\u50cf\uff0c\u7136\u540e\u5229\u7528\u56fe\u50cf\u5206\u5272\u6216\u56fe\u50cf\u4fee\u590d\u7b97\u6cd5\u6765\u8bc6\u522b\u548c\u53bb\u9664\u6c34\u5370\u3002\u5e38\u7528\u7684\u65b9\u6cd5\u5305\u62ec\u56fe\u50cf\u6a21\u7cca\u3001\u514b\u9686\u3001\u6216\u5229\u7528\u56fe\u50cf\u4fee\u590d\u529f\u80fd\uff0c\u4f8b\u5982OpenCV\u4e2d\u7684inpainting\u51fd\u6570\u3002\u5177\u4f53\u7684\u6b65\u9aa4\u5305\u62ec\u56fe\u50cf\u9884\u5904\u7406\u3001\u786e\u5b9a\u6c34\u5370\u533a\u57df\u3001\u5e94\u7528\u4fee\u590d\u7b97\u6cd5\u5e76\u6700\u7ec8\u4fdd\u5b58\u4fee\u6539\u540e\u7684\u56fe\u50cf\u3002<\/p>\n
\u53bb\u6c34\u5370\u65f6\u9700\u8981\u6ce8\u610f\u54ea\u4e9b\u6cd5\u5f8b\u548c\u9053\u5fb7\u95ee\u9898\uff1f<\/strong>
\u5728\u53bb\u9664\u6c34\u5370\u65f6\uff0c\u7528\u6237\u9700\u8c28\u614e\u8003\u8651\u7248\u6743\u548c\u4f7f\u7528\u6743\u95ee\u9898\u3002\u8bb8\u591a\u56fe\u50cf\u4e0a\u7684\u6c34\u5370\u662f\u4e3a\u4e86\u4fdd\u62a4\u7248\u6743\u800c\u6dfb\u52a0\u7684\uff0c\u53bb\u9664\u6c34\u5370\u53ef\u80fd\u4f1a\u4fb5\u72af\u7248\u6743\u3002\u56e0\u6b64\uff0c\u5728\u8fdb\u884c\u6c34\u5370\u53bb\u9664\u4e4b\u524d\uff0c\u5e94\u786e\u4fdd\u62e5\u6709\u76f8\u5e94\u7684\u6388\u6743\u6216\u8bb8\u53ef\uff0c\u6216\u8005\u786e\u4fdd\u8be5\u56fe\u50cf\u53ef\u4ee5\u5408\u6cd5\u4f7f\u7528\u3002\u540c\u65f6\uff0c\u5c0a\u91cd\u539f\u521b\u8005\u7684\u52b3\u52a8\u6210\u679c\u4e5f\u662f\u975e\u5e38\u91cd\u8981\u7684\u3002<\/p>\n
\u54ea\u4e9bPython\u5e93\u53ef\u4ee5\u4e0eOpenCV\u7ed3\u5408\u4f7f\u7528\u4ee5\u63d0\u9ad8\u53bb\u6c34\u5370\u7684\u6548\u679c\uff1f<\/strong>
\u9664\u4e86OpenCV\uff0c\u5176\u4ed6\u4e00\u4e9bPython\u5e93\u4e5f\u53ef\u4ee5\u5e2e\u52a9\u589e\u5f3a\u53bb\u6c34\u5370\u7684\u6548\u679c\u3002\u4f8b\u5982\uff0cNumPy\u53ef\u7528\u4e8e\u9ad8\u6548\u7684\u6570\u7ec4\u64cd\u4f5c\uff0cPillow\u53ef\u4ee5\u7528\u4e8e\u66f4\u590d\u6742\u7684\u56fe\u50cf\u5904\u7406\u548c\u683c\u5f0f\u8f6c\u6362\uff0cscikit-image\u63d0\u4f9b\u4e86\u4e00\u4e9b\u9ad8\u7ea7\u56fe\u50cf\u5904\u7406\u529f\u80fd\uff0c\u9002\u5408\u7528\u4e8e\u6c34\u5370\u53bb\u9664\u7684\u7b97\u6cd5\u5b9e\u73b0\u3002\u7ed3\u5408\u8fd9\u4e9b\u5e93\uff0c\u53ef\u4ee5\u5b9e\u73b0\u66f4\u52a0\u7075\u6d3b\u548c\u5f3a\u5927\u7684\u56fe\u50cf\u5904\u7406\u6548\u679c\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"Python OpenCV\u53bb\u9664\u6c34\u5370\u7684\u65b9\u6cd5\u6709\u591a\u79cd\uff0c\u5e38\u89c1\u7684\u5305\u62ec\u4f7f\u7528inpAInting\u6280\u672f\u3001\u56fe\u50cf\u4fee\u8865\u7b97\u6cd5\u3001\u8499\u7248\u5904\u7406 […]","protected":false},"author":3,"featured_media":1003114,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[37],"tags":[],"acf":[],"_links":{"self":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1003108"}],"collection":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/comments?post=1003108"}],"version-history":[{"count":"1","href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1003108\/revisions"}],"predecessor-version":[{"id":1003116,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1003108\/revisions\/1003116"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media\/1003114"}],"wp:attachment":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media?parent=1003108"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/categories?post=1003108"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/tags?post=1003108"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}