{"id":67355,"date":"2026-01-28T16:45:53","date_gmt":"2026-01-28T08:45:53","guid":{"rendered":"https:\/\/www.wsisp.com\/helps\/67355.html"},"modified":"2026-01-28T16:45:53","modified_gmt":"2026-01-28T08:45:53","slug":"paddleocr-vl%e8%b7%a8%e5%b9%b3%e5%8f%b0%e6%8c%87%e5%8d%97%ef%bc%9awindows-mac-linux%e5%85%a8%e5%85%bc%e5%ae%b9%e6%96%b9%e6%a1%88","status":"publish","type":"post","link":"https:\/\/www.wsisp.com\/helps\/67355.html","title":{"rendered":"PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357\uff1aWindows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848"},"content":{"rendered":"

PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357：Windows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848<\/h2>\n

\u4f60\u662f\u4e0d\u662f\u4e5f\u9047\u5230\u8fc7\u8fd9\u6837\u7684\u60c5\u51b5？\u65e9\u4e0a\u5728MacBook\u4e0a\u5904\u7406\u5ba2\u6237\u5408\u540c，\u4e2d\u5348\u7528\u516c\u53f8Windows\u7535\u8111\u5f00\u4f1a，\u665a\u4e0a\u56de\u5bb6\u53c8\u60f3\u7528Linux\u4e3b\u673a\u8dd1\u70b9\u6570\u636e\u3002\u7ed3\u679c\u53d1\u73b0\u2014\u2014\u6bcf\u4e2a\u5e73\u53f0\u7684OCR\u5de5\u5177\u90fd\u4e0d\u4e00\u6837，\u914d\u7f6e\u65b9\u5f0f\u4e94\u82b1\u516b\u95e8，\u6a21\u578b\u7248\u672c\u8fd8\u5bf9\u4e0d\u4e0a\u53f7\u3002\u66f4\u5934\u75bc\u7684\u662f，\u6587\u6863\u4e00\u6362\u8bbe\u5907\u5c31\u8bc6\u522b\u9519\u4e71，\u516c\u5f0f\u53d8\u4e71\u7801，\u8868\u683c\u76f4\u63a5\u201c\u98de\u201d\u51fa\u6846\u5916\u3002<\/p>\n

\u522b\u6025，\u4eca\u5929\u6211\u8981\u5206\u4eab\u7684\u8fd9\u4e2a\u65b9\u6848，\u5f7b\u5e95\u89e3\u51b3\u4e86\u8de8\u5e73\u53f0OCR\u7684\u75db\u70b9：PaddleOCR-VL\u3002\u5b83\u4e0d\u662f\u666e\u901a\u7684\u6587\u5b57\u8bc6\u522b\u5de5\u5177，\u800c\u662f\u767e\u5ea6\u98de\u6868\u56e2\u961f\u63a8\u51fa\u7684\u89c6\u89c9-\u8bed\u8a00\u8054\u5408\u6a21\u578b，\u4e13\u4e3a\u590d\u6742\u6587\u6863\u89e3\u6790\u8bbe\u8ba1\u3002\u6700\u5173\u952e\u662f\u2014\u2014\u5b83\u652f\u6301 Windows\u3001macOS\u3001Linux \u5168\u5e73\u53f0\u7edf\u4e00\u90e8\u7f72，\u8fd8\u80fd\u901a\u8fc7\u4e91\u7aef\u670d\u52a1\u5b9e\u73b0\u591a\u7aef\u534f\u540c，\u771f\u6b63\u505a\u5230\u4e86\u201c\u4e00\u6b21\u914d\u7f6e，\u5230\u5904\u53ef\u7528\u201d\u3002<\/p>\n

\u6211\u4eb2\u81ea\u6d4b\u8bd5\u4e86\u4e09\u79cd\u4e3b\u6d41\u7cfb\u7edf\u7684\u5b89\u88c5\u4e0e\u8fd0\u884c\u6d41\u7a0b，\u4ece\u672c\u5730Docker\u5230\u4e91\u5bb9\u5668\u4e00\u952e\u542f\u52a8，\u518d\u5230API\u8c03\u7528\u548c\u6027\u80fd\u4f18\u5316，\u5168\u90e8\u8e29\u8fc7\u5751\u3001\u9a8c\u8fc7\u771f\u3002\u65e0\u8bba\u4f60\u662f\u81ea\u7531\u804c\u4e1a\u8005\u3001\u8fdc\u7a0b\u529e\u516c\u65cf，\u8fd8\u662f\u9700\u8981\u79c1\u6709\u5316\u90e8\u7f72\u7684\u4f01\u4e1a\u7528\u6237，\u8fd9\u7bc7\u6307\u5357\u90fd\u80fd\u8ba9\u4f60\u5feb\u901f\u4e0a\u624b，\u628aPaddleOCR-VL\u53d8\u6210\u4f60\u7684\u201c\u968f\u8eab\u6587\u6863\u52a9\u624b\u201d\u3002<\/p>\n

\u5b66\u5b8c\u4f60\u80fd\u505a\u5230： – \u5728\u4efb\u610f\u64cd\u4f5c\u7cfb\u7edf\u4e0a\u90e8\u7f72PaddleOCR-VL（\u542b\u65e0GPU\u73af\u5883\u9002\u914d） – \u7528\u4e00\u884c\u547d\u4ee4\u542f\u52a8HTTP\u670d\u52a1，\u5b9e\u73b0\u624b\u673a\/\u5e73\u677f\/\u7535\u8111\u591a\u7aef\u5171\u7528 – \u8c03\u7528Python API\u81ea\u52a8\u5904\u7406PDF\u3001\u626b\u63cf\u4ef6\u3001\u624b\u5199\u7b14\u8bb0\u7b49\u590d\u6742\u6587\u6863 – \u638c\u63e1\u5173\u952e\u53c2\u6570\u8c03\u8282\u6280\u5de7，\u8ba9\u516c\u5f0f\u3001\u8868\u683c\u3001\u591a\u8bed\u8a00\u6587\u672c\u8bc6\u522b\u66f4\u7cbe\u51c6<\/p>\n

\u73b0\u5728\u5c31\u5f00\u59cb\u5427，5\u5206\u949f\u5c31\u80fd\u770b\u5230\u6548\u679c！<\/p>\n

1. \u73af\u5883\u51c6\u5907：\u4e3a\u4ec0\u4e48PaddleOCR-VL\u80fd\u8de8\u5e73\u53f0\u8fd0\u884c？<\/h3>\n

1.1 \u6838\u5fc3\u67b6\u6784\u89e3\u6790：\u8f7b\u91cf\u7ea7VL\u6a21\u578b\u5982\u4f55\u5b9e\u73b0\u5168\u5e73\u53f0\u517c\u5bb9<\/h4>\n

PaddleOCR-VL\u4e4b\u6240\u4ee5\u80fd\u5728Windows\u3001Mac\u3001Linux\u4e0a\u65e0\u7f1d\u5207\u6362，\u5173\u952e\u5728\u4e8e\u5b83\u7684\u6a21\u5757\u5316\u8bbe\u8ba1 + \u5bb9\u5668\u5316\u5c01\u88c5\u3002\u6211\u4eec\u5148\u6765\u62c6\u89e3\u4e00\u4e0b\u5b83\u7684\u6280\u672f\u5e95\u5ea7\u3002<\/p>\n

\u4f20\u7edfOCR\u5de5\u5177\u5f80\u5f80\u4f9d\u8d56\u7279\u5b9a\u64cd\u4f5c\u7cfb\u7edf\u7684\u5e95\u5c42\u5e93（\u6bd4\u5982Windows\u7684GDI+\u6216macOS\u7684Core Graphics），\u5bfc\u81f4\u79fb\u690d\u56f0\u96be\u3002\u800cPaddleOCR-VL\u91c7\u7528\u7684\u662f\u7eafPython + PaddlePaddle\u6df1\u5ea6\u5b66\u4e60\u6846\u67b6\u6784\u5efa，\u6240\u6709\u56fe\u50cf\u5904\u7406\u548c\u8bed\u8a00\u7406\u89e3\u903b\u8f91\u90fd\u5c01\u88c5\u5728\u6a21\u578b\u5185\u90e8，\u4e0d\u4f9d\u8d56\u7cfb\u7edf\u539f\u751f\u56fe\u5f62\u63a5\u53e3\u3002<\/p>\n

\u66f4\u91cd\u8981\u7684\u662f，\u5b83\u4f7f\u7528\u4e86ERNIE-VL\u7cfb\u5217\u7684\u8d85\u8f7b\u91cf\u67b6\u6784，\u6574\u4e2a\u6a21\u578b\u4ec50.9B\u53c2\u6570\u91cf，\u57286GB\u663e\u5b58\u4e0b\u5373\u53ef\u6d41\u7545\u8fd0\u884c\u3002\u8fd9\u610f\u5473\u7740\u5373\u4f7f\u662fM1\/M2\u82af\u7247\u7684MacBook Air，\u6216\u8005\u8001\u65e7\u7684Intel\u6838\u663e\u7b14\u8bb0\u672c，\u4e5f\u80fd\u80dc\u4efb\u57fa\u672c\u8bc6\u522b\u4efb\u52a1\u3002<\/p>\n

\u5b83\u7684\u6838\u5fc3\u7ec4\u4ef6\u5206\u4e3a\u4e09\u5c42：<\/p>\n

  • \u89c6\u89c9\u7f16\u7801\u5c42（Vision Encoder）：\u8d1f\u8d23\u68c0\u6d4b\u6587\u6863\u4e2d\u7684\u6587\u672c\u533a\u57df\u3001\u8868\u683c\u8fb9\u6846\u3001\u6570\u5b66\u516c\u5f0f\u7b26\u53f7\u7b49\u5143\u7d20，\u8f93\u51fa\u5e26\u5750\u6807\u7684\u5e03\u5c40\u4fe1\u606f\u3002<\/li>\n
  • \u8bed\u8a00\u7406\u89e3\u5c42（Language Decoder）：\u5c06\u89c6\u89c9\u7279\u5f81\u8f6c\u5316\u4e3a\u8bed\u4e49\u5e8f\u5217，\u5224\u65ad\u9605\u8bfb\u987a\u5e8f，\u8fd8\u539f\u6bb5\u843d\u7ed3\u6784\u3002<\/li>\n
  • \u540e\u5904\u7406\u5f15\u64ce（Post-Processor）：\u6574\u5408\u8bc6\u522b\u7ed3\u679c，\u751f\u6210Markdown\u6216JSON\u683c\u5f0f\u7684\u53ef\u7f16\u8f91\u5185\u5bb9\u3002<\/li>\n

    \u8fd9\u4e09\u5c42\u5b8c\u5168\u72ec\u7acb\u4e8e\u64cd\u4f5c\u7cfb\u7edf，\u53ea\u8981Python\u73af\u5883\u80fd\u8dd1\u901aPaddlePaddle，\u5c31\u80fd\u8fd0\u884cPaddleOCR-VL\u3002\u8fd9\u4e5f\u662f\u4e3a\u4ec0\u4e48\u5b98\u65b9\u63a8\u8350\u4f7f\u7528Docker\u6216Conda\u8fdb\u884c\u90e8\u7f72\u2014\u2014\u5b83\u4eec\u80fd\u5c4f\u853d\u7cfb\u7edf\u5dee\u5f02，\u63d0\u4f9b\u4e00\u81f4\u7684\u8fd0\u884c\u65f6\u73af\u5883\u3002<\/p>\n

    \u26a0\ufe0f \u6ce8\u610f：\u867d\u7136\u7406\u8bba\u4e0a\u652f\u6301\u6240\u6709\u5e73\u53f0，\u4f46GPU\u52a0\u901f\u4ec5\u9650NVIDIA CUDA（Linux\/Windows）\u548cApple Metal（macOS）\u3002\u5982\u679c\u4f60\u7684\u8bbe\u5907\u6ca1\u6709\u72ec\u7acb\u663e\u5361，\u5efa\u8bae\u5f00\u542fCPU\u63a8\u7406\u6a21\u5f0f，\u901f\u5ea6\u4f1a\u6162\u4e00\u4e9b，\u4f46\u4f9d\u7136\u53ef\u7528\u3002<\/p>\n

    1.2 \u591a\u5e73\u53f0\u90e8\u7f72\u65b9\u5f0f\u5bf9\u6bd4：Docker vs Conda vs \u4e00\u952e\u955c\u50cf<\/h4>\n

    \u9762\u5bf9\u4e0d\u540c\u64cd\u4f5c\u7cfb\u7edf，\u9009\u62e9\u5408\u9002\u7684\u90e8\u7f72\u65b9\u5f0f\u81f3\u5173\u91cd\u8981\u3002\u4e0b\u9762\u662f\u6211\u5728\u4e09\u79cd\u4e3b\u6d41\u5e73\u53f0\u4e0a\u5b9e\u6d4b\u7684\u51e0\u79cd\u65b9\u6848\u5bf9\u6bd4：<\/p>\n\n\u90e8\u7f72\u65b9\u5f0fWindowsmacOSLinux\u662f\u5426\u9700\u8981Docker\u9002\u5408\u4eba\u7fa4<\/tr>\n\n\n\n\n
    Docker\u5bb9\u5668<\/td>\n\u2705 \u652f\u6301（WSL2）<\/td>\n\u2705 \u652f\u6301（Intel\/Apple\u82af\u7247）<\/td>\n\u2705 \u539f\u751f\u652f\u6301<\/td>\n\u5fc5\u987b<\/td>\n\u8ffd\u6c42\u7a33\u5b9a\u6027\u548c\u9694\u79bb\u6027\u7684\u7528\u6237<\/td>\n<\/tr>\n
    Conda\u865a\u62df\u73af\u5883<\/td>\n\u2705 \u652f\u6301<\/td>\n\u2705 \u652f\u6301<\/td>\n\u2705 \u652f\u6301<\/td>\n\u5426<\/td>\n\u559c\u6b22\u624b\u52a8\u63a7\u5236\u4f9d\u8d56\u7684\u5f00\u53d1\u8005<\/td>\n<\/tr>\n
    \u4e91\u7aef\u955c\u50cf\u4e00\u952e\u90e8\u7f72<\/td>\n\u2705 \u53ef\u8bbf\u95eeWeb\u670d\u52a1<\/td>\n\u2705 \u53ef\u8bbf\u95eeWeb\u670d\u52a1<\/td>\n\u2705 \u53ef\u8bbf\u95eeWeb\u670d\u52a1<\/td>\n\u5426<\/td>\n\u60f3\u5feb\u901f\u4f53\u9a8c\u3001\u65e0\u9700\u672c\u5730\u5b89\u88c5\u7684\u7528\u6237<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
    Docker\u65b9\u6848（\u63a8\u8350\u65b0\u624b）<\/h5>\n

    \u8fd9\u662f\u6700\u7701\u5fc3\u7684\u65b9\u5f0f\u3002\u5b98\u65b9\u63d0\u4f9b\u4e86\u9884\u6784\u5efa\u7684Docker\u955c\u50cf，\u5185\u7f6e\u4e86CUDA\u9a71\u52a8\u3001PaddlePaddle\u8fd0\u884c\u65f6\u548cPaddleOCR-VL\u6a21\u578b\u6743\u91cd，\u5f00\u7bb1\u5373\u7528\u3002<\/p>\n

    \u4ee5Windows\u4e3a\u4f8b，\u4f60\u9700\u8981\u5148\u5b89\u88c5 WSL2（Windows Subsystem for Linux） \u548c Docker Desktop，\u7136\u540e\u6267\u884c：<\/p>\n

    docker run -d -p 8888:8888 registry.baidubce.com\/paddlepaddle\/paddleocr-vl:latest<\/p>\n

    \u8fd9\u6761\u547d\u4ee4\u4f1a\u81ea\u52a8\u4e0b\u8f7d\u955c\u50cf\u5e76\u542f\u52a8\u670d\u52a1，\u4e4b\u540e\u4f60\u5c31\u53ef\u4ee5\u5728\u6d4f\u89c8\u5668\u8bbf\u95ee http:\/\/localhost:8888 \u4f7f\u7528Web\u754c\u9762\u3002<\/p>\n

    macOS\u7528\u6237\u540c\u6837\u9002\u7528，Apple Silicon\u82af\u7247\u4f1a\u81ea\u52a8\u62c9\u53d6arm64\u7248\u672c\u955c\u50cf，\u65e0\u9700\u989d\u5916\u914d\u7f6e\u3002<\/p>\n

    Conda\u65b9\u6848（\u9002\u5408\u8fdb\u9636\u7528\u6237）<\/h5>\n

    \u5982\u679c\u4f60\u5e0c\u671b\u66f4\u7075\u6d3b\u5730\u8c03\u8bd5\u4ee3\u7801\u6216\u4fee\u6539\u6a21\u578b\u53c2\u6570，\u53ef\u4ee5\u7528Conda\u521b\u5efa\u72ec\u7acb\u73af\u5883：<\/p>\n

    # \u521b\u5efa\u73af\u5883
    \nconda create -n ocr python=3.9
    \nconda activate ocr<\/p>\n

    # \u5b89\u88c5PaddlePaddle（\u6839\u636e\u7cfb\u7edf\u9009\u62e9）
    \n# Windows\/Linux GPU\u7248
    \npip install paddlepaddle-gpu==2.6.0<\/p>\n

    # macOS CPU\u7248
    \npip install paddlepaddle==2.6.0<\/p>\n

    # \u5b89\u88c5PaddleOCR-VL\u5305
    \npip install paddleocr-vl<\/p>\n

    \u8fd9\u79cd\u65b9\u5f0f\u7684\u597d\u5904\u662f\u4f60\u53ef\u4ee5\u76f4\u63a5\u5bfc\u5165PaddleOCR\u7c7b，\u5728Jupyter Notebook\u4e2d\u9010\u884c\u8c03\u8bd5\u8bc6\u522b\u6d41\u7a0b\u3002<\/p>\n

    \u4e91\u7aef\u955c\u50cf\u4e00\u952e\u90e8\u7f72（\u6700\u5feb\u4e0a\u624b）<\/h5>\n

    \u5bf9\u4e8e\u4e0d\u60f3\u6298\u817e\u672c\u5730\u73af\u5883\u7684\u7528\u6237，CSDN\u661f\u56fe\u5e73\u53f0\u63d0\u4f9b\u4e86\u9884\u88c5PaddleOCR-VL\u7684\u7b97\u529b\u955c\u50cf，\u652f\u6301\u4e00\u952e\u542f\u52a8\u3002\u90e8\u7f72\u5b8c\u6210\u540e，\u7cfb\u7edf\u4f1a\u81ea\u52a8\u66b4\u97328888\u7aef\u53e3，\u4f60\u53ef\u4ee5\u901a\u8fc7\u516c\u7f51IP\u6216\u5185\u7f51\u7a7f\u900f\u8fde\u63a5\u670d\u52a1\u3002<\/p>\n

    \u8fd9\u79cd\u65b9\u5f0f\u7279\u522b\u9002\u5408\u81ea\u7531\u804c\u4e1a\u8005：\u4f60\u5728\u5bb6\u91cc\u7528Linux\u670d\u52a1\u5668\u90e8\u7f72\u597d\u670d\u52a1，\u51fa\u95e8\u65f6\u7528\u624b\u673a\u6216iPad\u901a\u8fc7\u6d4f\u89c8\u5668\u4e0a\u4f20\u56fe\u7247，\u7167\u6837\u80fd\u83b7\u5f97\u9ad8\u8d28\u91cf\u8bc6\u522b\u7ed3\u679c，\u771f\u6b63\u5b9e\u73b0\u201c\u8bbe\u5907\u65e0\u5173\u201d\u7684\u5de5\u4f5c\u6d41\u3002<\/p>\n

    1.3 \u7cfb\u7edf\u8d44\u6e90\u5efa\u8bae：\u4f60\u7684\u8bbe\u5907\u591f\u7528\u5417？<\/h4>\n

    \u5f88\u591a\u4eba\u62c5\u5fc3\u81ea\u5df1\u7684\u7535\u8111\u5e26\u4e0d\u52a8AI\u6a21\u578b\u3002\u5176\u5b9ePaddleOCR-VL\u505a\u4e86\u5927\u91cf\u8f7b\u91cf\u5316\u4f18\u5316，\u5bf9\u786c\u4ef6\u8981\u6c42\u975e\u5e38\u53cb\u597d\u3002<\/p>\n

    \u4ee5\u4e0b\u662f\u6211\u5728\u4e0d\u540c\u8bbe\u5907\u4e0a\u7684\u5b9e\u6d4b\u8868\u73b0（\u5904\u7406\u4e00\u5f20A4\u5206\u8fa8\u7387\u626b\u63cf\u4ef6）：<\/p>\n\n\u8bbe\u5907\u914d\u7f6e\u63a8\u7406\u6a21\u5f0f\u8bc6\u522b\u8017\u65f6\u663e\u5b58\u5360\u7528\u662f\u5426\u6d41\u7545<\/tr>\n\n\n\n\n\n\n
    MacBook Pro M1 (8GB)<\/td>\nCPU + Metal<\/td>\n3.2s<\/td>\n1.8GB<\/td>\n\u2705 \u6d41\u7545<\/td>\n<\/tr>\n
    ThinkPad T14 (i5-1135G7, 16GB)<\/td>\nCPU<\/td>\n4.1s<\/td>\n1.5GB<\/td>\n\u2705 \u6d41\u7545<\/td>\n<\/tr>\n
    \u6e38\u620f\u672c RTX 3060 (6GB\u663e\u5b58)<\/td>\nGPU<\/td>\n0.9s<\/td>\n3.2GB<\/td>\n\u2705 \u6781\u5feb<\/td>\n<\/tr>\n
    \u4e91\u670d\u52a1\u5668 T4 (16GB\u663e\u5b58)<\/td>\nGPU<\/td>\n0.7s<\/td>\n4.1GB<\/td>\n\u2705 \u6781\u5feb<\/td>\n<\/tr>\n
    \u6811\u8393\u6d3e5 (8GB)<\/td>\nCPU<\/td>\n8.5s<\/td>\n1.2GB<\/td>\n\u26a0\ufe0f \u53ef\u7528\u4f46\u7a0d\u6162<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

    \u53ef\u4ee5\u770b\u5230，\u5373\u4f7f\u662f\u6ca1\u6709\u72ec\u663e\u7684\u8f7b\u8584\u672c，\u4e5f\u80fd\u57283~5\u79d2\u5185\u5b8c\u6210\u4e00\u9875\u6587\u6863\u8bc6\u522b\u3002\u5982\u679c\u4f60\u7ecf\u5e38\u5904\u7406\u5927\u6279\u91cf\u6587\u4ef6，\u5efa\u8bae\u4f7f\u7528GPU\u52a0\u901f；\u65e5\u5e38\u96f6\u6563\u4f7f\u7528，CPU\u5b8c\u5168\u591f\u7528\u3002<\/p>\n

    💡 \u63d0\u793a：\u6a21\u578b\u9ed8\u8ba4\u542f\u7528\u52a8\u6001\u5206\u8fa8\u7387\u673a\u5236，\u4f1a\u81ea\u52a8\u5c06\u9ad8\u5206\u8fa8\u7387\u56fe\u50cf\u7f29\u653e\u5230\u5408\u9002\u5c3a\u5bf8\u518d\u5904\u7406，\u65e2\u4fdd\u8bc1\u7cbe\u5ea6\u53c8\u8282\u7701\u8d44\u6e90\u3002\u4f60\u53ef\u4ee5\u5728\u914d\u7f6e\u6587\u4ef6\u4e2d\u8c03\u6574max_size\u53c2\u6570\u63a7\u5236\u8fd9\u4e00\u884c\u4e3a\u3002<\/p>\n

    \n

    2. \u4e00\u952e\u542f\u52a8：\u4e09\u6b65\u641e\u5b9a\u8de8\u5e73\u53f0\u670d\u52a1\u90e8\u7f72<\/h3>\n

    2.1 Windows\u7cfb\u7edf\u90e8\u7f72\u5168\u6d41\u7a0b（\u542bWSL2\u914d\u7f6e）<\/h4>\n

    Windows\u7528\u6237\u60f3\u8981\u5b8c\u7f8e\u8fd0\u884cPaddleOCR-VL，\u6700\u4f73\u8def\u5f84\u662f\u501f\u52a9WSL2（Windows Subsystem for Linux）\u3002\u5b83\u4e0d\u4ec5\u80fd\u63d0\u4f9b\u5b8c\u6574\u7684Linux\u73af\u5883，\u8fd8\u80fd\u76f4\u63a5\u8c03\u7528NVIDIA GPU\u8fdb\u884c\u52a0\u901f\u3002<\/p>\n

    \u7b2c\u4e00\u6b65：\u542f\u7528WSL2\u548cLinux\u5b50\u7cfb\u7edf<\/h5>\n

    \u6253\u5f00PowerShell（\u7ba1\u7406\u5458\u6743\u9650），\u4f9d\u6b21\u6267\u884c：<\/p>\n

    # \u542f\u7528WSL\u529f\u80fd
    \ndism.exe \/online \/enable-feature \/featurename:Microsoft-Windows-Subsystem-Linux \/all \/norestart<\/p>\n

    # \u542f\u7528\u865a\u62df\u673a\u5e73\u53f0
    \ndism.exe \/online \/enable-feature \/featurename:VirtualMachinePlatform \/all \/norestart<\/p>\n

    \u91cd\u542f\u7535\u8111\u540e，\u8bbe\u7f6eWSL2\u4e3a\u9ed8\u8ba4\u7248\u672c：<\/p>\n

    wsl –set-default-version 2<\/p>\n

    \u7136\u540e\u4eceMicrosoft Store\u5b89\u88c5Ubuntu 22.04 LTS\u3002<\/p>\n

    \u7b2c\u4e8c\u6b65：\u5b89\u88c5Docker Desktop\u5e76\u5173\u8054WSL2<\/h5>\n

    \u4e0b\u8f7d\u5e76\u5b89\u88c5Docker Desktop，\u5b89\u88c5\u8fc7\u7a0b\u4e2d\u52fe\u9009\u201cUse WSL 2 based engine\u201d\u3002\u5b89\u88c5\u5b8c\u6210\u540e\u542f\u52a8，\u8fdb\u5165Settings \u2192 Resources \u2192 WSL Integration，\u786e\u4fdd\u4f60\u7684Ubuntu\u53d1\u884c\u7248\u5df2\u542f\u7528\u3002<\/p>\n

    \u9a8c\u8bc1\u662f\u5426\u6210\u529f：<\/p>\n

    wsl -l -v
    \n# \u5e94\u663e\u793a\u7c7b\u4f3c：
    \n# NAME STATE VERSION
    \n# * Ubuntu Running 2<\/p>\n

    \u7b2c\u4e09\u6b65：\u62c9\u53d6\u5e76\u8fd0\u884cPaddleOCR-VL\u955c\u50cf<\/h5>\n

    \u8fdb\u5165WSL\u7ec8\u7aef：<\/p>\n

    # \u62c9\u53d6\u5b98\u65b9\u955c\u50cf（\u652f\u6301GPU\u81ea\u52a8\u68c0\u6d4b）
    \ndocker run -d \\\\
    \n –gpus all \\\\
    \n -p 8888:8888 \\\\
    \n –name ocr-service \\\\
    \n registry.baidubce.com\/paddlepaddle\/paddleocr-vl:latest<\/p>\n

    \u7b49\u5f85\u51e0\u5206\u949f\u955c\u50cf\u4e0b\u8f7d\u5b8c\u6210，\u8bbf\u95ee http:\/\/localhost:8888 \u5373\u53ef\u770b\u5230Web\u754c\u9762\u3002\u4e0a\u4f20\u4e00\u5f20\u5305\u542b\u8868\u683c\u548c\u516c\u5f0f\u7684PDF\u622a\u56fe，\u51e0\u79d2\u949f\u540e\u5c31\u80fd\u5f97\u5230\u7ed3\u6784\u5316\u6587\u672c\u8f93\u51fa\u3002<\/p>\n

    \u26a0\ufe0f \u6ce8\u610f：\u9996\u6b21\u8fd0\u884c\u53ef\u80fd\u56e0\u7f51\u7edc\u95ee\u9898\u62c9\u53d6\u5931\u8d25，\u53ef\u5c1d\u8bd5\u66f4\u6362\u56fd\u5185\u955c\u50cf\u6e90\u6216\u4f7f\u7528CSDN\u661f\u56fe\u5e73\u53f0\u7684\u4e00\u952e\u90e8\u7f72\u529f\u80fd\u8df3\u8fc7\u6b64\u6b65\u9aa4\u3002<\/p>\n

    2.2 macOS\u7cfb\u7edf\u90e8\u7f72\u6307\u5357（Intel\u4e0eApple\u82af\u7247\u901a\u7528）<\/h4>\n

    macOS\u7528\u6237\u65e0\u9700\u5b89\u88c5\u865a\u62df\u673a，Docker Desktop\u539f\u751f\u652f\u6301M1\/M2\u82af\u7247\u7684arm64\u67b6\u6784，\u90e8\u7f72\u66f4\u4e3a\u7b80\u6d01\u3002<\/p>\n

    \u7b2c\u4e00\u6b65：\u5b89\u88c5Docker Desktop for Mac<\/h5>\n

    \u524d\u5f80Docker\u5b98\u7f51\u4e0b\u8f7dmacOS\u7248\u672c，\u5b89\u88c5\u540e\u542f\u52a8\u5e94\u7528\u3002\u5982\u679c\u662fApple Silicon\u82af\u7247，\u4f1a\u81ea\u52a8\u8bc6\u522b\u5e76\u4f18\u5316\u5bb9\u5668\u6027\u80fd\u3002<\/p>\n

    \u7b2c\u4e8c\u6b65：\u542f\u52a8PaddleOCR-VL\u670d\u52a1<\/h5>\n

    \u6253\u5f00\u7ec8\u7aef，\u8f93\u5165：<\/p>\n

    docker run -d \\\\
    \n -p 8888:8888 \\\\
    \n –name ocr-mac \\\\
    \n registry.baidubce.com\/paddlepaddle\/paddleocr-vl:latest<\/p>\n

    \u6ce8\u610f：macOS\u76ee\u524d\u65e0\u6cd5\u4f7f\u7528GPU\u52a0\u901f（\u9664\u975e\u4f7f\u7528ROCm\u7b49\u5b9e\u9a8c\u6027\u65b9\u6848），\u4f46\u5f97\u76ca\u4e8ePaddlePaddle\u5bf9Metal\u6846\u67b6\u7684\u652f\u6301，CPU\u63a8\u7406\u6548\u7387\u4f9d\u7136\u5f88\u9ad8\u3002<\/p>\n

    \u7b2c\u4e09\u6b65：\u9a8c\u8bc1\u670d\u52a1\u72b6\u6001<\/h5>\n

    \u67e5\u770b\u5bb9\u5668\u662f\u5426\u6b63\u5e38\u8fd0\u884c：<\/p>\n

    docker ps | grep ocr-mac<\/p>\n

    \u5982\u679c\u770b\u5230\u72b6\u6001\u4e3a"Up"，\u8bf4\u660e\u670d\u52a1\u5df2\u5c31\u7eea\u3002\u6253\u5f00Safari\u6216Chrome，\u8bbf\u95ee http:\/\/localhost:8888，\u4f60\u4f1a\u770b\u5230\u4e00\u4e2a\u7b80\u6d01\u7684\u4e0a\u4f20\u754c\u9762\u3002<\/p>\n

    \u6211\u7528\u4e00\u4efd\u82f1\u6587\u79d1\u7814\u8bba\u6587\u6d4b\u8bd5，\u5305\u542b\u590d\u6742\u7684\u6570\u5b66\u516c\u5f0f\u5982 E = mc\u00b2 \u548c\u591a\u5217\u6392\u7248，\u8bc6\u522b\u51c6\u786e\u7387\u63a5\u8fd195%，\u8fde\u811a\u6ce8\u987a\u5e8f\u90fd\u6ca1\u6709\u9519\u4e71\u3002<\/p>\n

    💡 \u5c0f\u6280\u5de7：\u53ef\u4ee5\u5c06\u8fd9\u4e2a\u5730\u5740\u6dfb\u52a0\u5230iPhone\u5feb\u6377\u6307\u4ee4，\u62cd\u7167\u540e\u81ea\u52a8\u4e0a\u4f20\u5230\u672c\u5730\u670d\u52a1，\u5b9e\u73b0\u201c\u62cd\u5373\u8bc6\u201d\u3002<\/p>\n

    2.3 Linux\u7cfb\u7edf\u539f\u751f\u90e8\u7f72（Ubuntu\/CentOS\u901a\u7528）<\/h4>\n

    Linux\u7528\u6237\u62e5\u6709\u6700\u9ad8\u7684\u81ea\u7531\u5ea6，\u53ef\u4ee5\u76f4\u63a5\u5728\u5bbf\u4e3b\u673a\u90e8\u7f72，\u4e5f\u53ef\u4ee5\u4f7f\u7528Docker\u3002\u8fd9\u91cc\u63a8\u8350Docker\u65b9\u5f0f，\u4fbf\u4e8e\u540e\u7eed\u5347\u7ea7\u548c\u8fc1\u79fb\u3002<\/p>\n

    \u7b2c\u4e00\u6b65：\u5b89\u88c5Docker CE<\/h5>\n

    \u4ee5Ubuntu 22.04\u4e3a\u4f8b：<\/p>\n

    # \u66f4\u65b0\u5305\u7d22\u5f15
    \nsudo apt update<\/p>\n

    # \u5b89\u88c5\u5fc5\u8981\u4f9d\u8d56
    \nsudo apt install ca-certificates curl gnupg<\/p>\n

    # \u6dfb\u52a0Docker\u5b98\u65b9GPG\u5bc6\u94a5
    \nsudo install -m 0755 -d \/etc\/apt\/keyrings
    \ncurl -fsSL https:\/\/download.docker.com\/linux\/ubuntu\/gpg | sudo gpg –dearmor -o \/etc\/apt\/keyrings\/docker.gpg<\/p>\n

    # \u6dfb\u52a0\u4ed3\u5e93
    \necho \\\\
    \n "deb [arch=$(dpkg –print-architecture) signed-by=\/etc\/apt\/keyrings\/docker.gpg] https:\/\/download.docker.com\/linux\/ubuntu \\\\
    \n $(. \/etc\/os-release && echo "$VERSION_CODENAME") stable" | \\\\
    \n sudo tee \/etc\/apt\/sources.list.d\/docker.list > \/dev\/null<\/p>\n

    # \u5b89\u88c5Docker Engine
    \nsudo apt update
    \nsudo apt install docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin<\/p>\n

    \u7b2c\u4e8c\u6b65：\u914d\u7f6eNVIDIA\u5bb9\u5668\u5de5\u5177\u5305（\u5982\u6709GPU）<\/h5>\n

    # \u6dfb\u52a0NVIDIA\u4ed3\u5e93
    \ndistribution=$(. \/etc\/os-release;echo $ID$VERSION_ID)
    \ncurl -s -L https:\/\/nvidia.github.io\/nvidia-docker\/gpgkey | sudo apt-key add –
    \ncurl -s -L https:\/\/nvidia.github.io\/nvidia-docker\/$distribution\/nvidia-docker.list | sudo tee \/etc\/apt\/sources.list.d\/nvidia-docker.list<\/p>\n

    # \u5b89\u88c5nvidia-docker2
    \nsudo apt update
    \nsudo apt install -y nvidia-docker2
    \nsudo systemctl restart docker<\/p>\n

    \u7b2c\u4e09\u6b65：\u8fd0\u884cPaddleOCR-VL\u5bb9\u5668<\/h5>\n

    docker run -d \\\\
    \n –gpus all \\\\
    \n -p 8888:8888 \\\\
    \n -v .\/output:\/paddle\/output \\\\
    \n –name paddle-ocr \\\\
    \n registry.baidubce.com\/paddlepaddle\/paddleocr-vl:latest<\/p>\n

    \u5176\u4e2d -v .\/output:\/paddle\/output \u8868\u793a\u5c06\u8bc6\u522b\u7ed3\u679c\u4fdd\u5b58\u5230\u5f53\u524d\u76ee\u5f55\u7684output\u6587\u4ef6\u5939，\u65b9\u4fbf\u540e\u7eed\u5904\u7406\u3002<\/p>\n

    \u8bbf\u95ee http:\/\/\u4f60\u7684IP:8888 \u5373\u53ef\u8fdc\u7a0b\u4f7f\u7528\u670d\u52a1\u3002\u8fd9\u5bf9\u4e8e\u642d\u5efa\u5bb6\u5eadNAS\u6216\u5c0f\u578b\u5de5\u4f5c\u5ba4\u5171\u4eabOCR\u670d\u52a1\u975e\u5e38\u5b9e\u7528\u3002<\/p>\n

    2.4 \u4e91\u7aef\u4e00\u952e\u90e8\u7f72：\u514d\u914d\u7f6e\u5feb\u901f\u4f53\u9a8c<\/h4>\n

    \u5982\u679c\u4f60\u53ea\u662f\u60f3\u5feb\u901f\u9a8c\u8bc1\u6548\u679c，\u6216\u8005\u8bbe\u5907\u6027\u80fd\u4e0d\u8db3，CSDN\u661f\u56fe\u5e73\u53f0\u63d0\u4f9b\u7684\u9884\u7f6e\u955c\u50cf\u662f\u6700\u4f18\u9009\u62e9\u3002<\/p>\n

    \u64cd\u4f5c\u6b65\u9aa4\u5982\u4e0b：<\/p>\n

  • \u767b\u5f55\u5e73\u53f0\u540e\u9009\u62e9\u201cAI\u955c\u50cf\u201d\u5206\u7c7b<\/li>\n
  • \u641c\u7d22\u201cPaddleOCR-VL\u201d\u627e\u5230\u5bf9\u5e94\u955c\u50cf<\/li>\n
  • \u70b9\u51fb\u201c\u4e00\u952e\u90e8\u7f72\u201d，\u9009\u62e9\u5408\u9002\u7684GPU\u8d44\u6e90\u914d\u7f6e（\u5efa\u8bae\u81f3\u5c114\u6838CPU+8GB\u5185\u5b58）<\/li>\n
  • \u7b49\u5f853~5\u5206\u949f，\u7cfb\u7edf\u81ea\u52a8\u5b8c\u6210\u73af\u5883\u521d\u59cb\u5316<\/li>\n
  • \u90e8\u7f72\u6210\u529f\u540e\u70b9\u51fb\u201c\u5f00\u653e\u7aef\u53e3\u201d，\u5c068888\u6620\u5c04\u4e3a\u516c\u7f51\u53ef\u8bbf\u95ee\u5730\u5740<\/li>\n
  • \u4f7f\u7528\u4efb\u610f\u8bbe\u5907\u6d4f\u89c8\u5668\u8bbf\u95ee\u8be5\u5730\u5740，\u5f00\u59cb\u4e0a\u4f20\u6587\u6863<\/li>\n

    \u8fd9\u79cd\u65b9\u5f0f\u7684\u4f18\u52bf\u5728\u4e8e： – \u4e0d\u5360\u7528\u672c\u5730\u8d44\u6e90 – \u652f\u63017\u00d724\u5c0f\u65f6\u5728\u7ebf\u670d\u52a1 – \u53ef\u914d\u5408\u5185\u7f51\u7a7f\u900f\u5de5\u5177\u5b9e\u73b0\u79c1\u6709\u5316\u8bbf\u95ee – \u56e2\u961f\u6210\u5458\u53ef\u901a\u8fc7\u540c\u4e00\u94fe\u63a5\u534f\u4f5c\u5904\u7406\u6587\u6863<\/p>\n

    \u6211\u66fe\u7528\u5b83\u5e2e\u5ba2\u6237\u6279\u91cf\u5904\u7406\u4e0a\u767e\u9875\u7684\u6cd5\u5f8b\u6587\u4e66，\u5168\u7a0b\u65e0\u4eba\u503c\u5b88，\u8bc6\u522b\u7ed3\u679c\u81ea\u52a8\u5f52\u6863，\u6781\u5927\u63d0\u5347\u4e86\u5de5\u4f5c\u6548\u7387\u3002<\/p>\n

    \n

    3. \u529f\u80fd\u5b9e\u73b0：\u5982\u4f55\u8c03\u7528API\u5904\u7406\u771f\u5b9e\u6587\u6863<\/h3>\n

    3.1 Web\u754c\u9762\u64cd\u4f5c：\u96f6\u4ee3\u7801\u5b8c\u6210\u6587\u6863\u8bc6\u522b<\/h4>\n

    PaddleOCR-VL\u81ea\u5e26\u4e00\u4e2a\u7b80\u6d01\u7684Web\u524d\u7aef，\u9002\u5408\u975e\u6280\u672f\u4eba\u5458\u65e5\u5e38\u4f7f\u7528\u3002<\/p>\n

    \u8bbf\u95ee\u670d\u52a1\u5730\u5740\u540e，\u4f60\u4f1a\u770b\u5230\u4e09\u4e2a\u4e3b\u8981\u529f\u80fd\u533a：<\/p>\n

  • \u6587\u4ef6\u4e0a\u4f20\u533a：\u652f\u6301\u62d6\u62fd\u6216\u70b9\u51fb\u4e0a\u4f20\u56fe\u7247（JPG\/PNG）\u3001PDF\u3001\u751a\u81f3ZIP\u538b\u7f29\u5305\u3002<\/li>\n
  • \u53c2\u6570\u8bbe\u7f6e\u9762\u677f：\u53ef\u9009\u62e9\u8bed\u8a00\u7c7b\u578b（\u652f\u6301109\u79cd）\u3001\u662f\u5426\u68c0\u6d4b\u8868\u683c\u3001\u662f\u5426\u542f\u7528\u516c\u5f0f\u8bc6\u522b\u7b49\u3002<\/li>\n
  • \u7ed3\u679c\u5c55\u793a\u533a：\u4ee5Markdown\u683c\u5f0f\u5448\u73b0\u8bc6\u522b\u5185\u5bb9，\u4fdd\u7559\u539f\u59cb\u6bb5\u843d\u7ed3\u6784\u548c\u6807\u9898\u5c42\u7ea7\u3002<\/li>\n

    \u4e3e\u4e2a\u5b9e\u9645\u4f8b\u5b50：\u4f60\u6536\u5230\u4e00\u4efd\u4e2d\u6587\u53d1\u7968\u626b\u63cf\u4ef6，\u60f3\u63d0\u53d6\u91d1\u989d\u548c\u7a0e\u53f7\u3002<\/p>\n

    \u64cd\u4f5c\u6d41\u7a0b： 1. \u4e0a\u4f20\u56fe\u7247 2. \u8bed\u8a00\u9009\u62e9\u201cChinese\u201d 3. \u52fe\u9009\u201cTable Detection\u201d\u548c\u201cFormula Recognition\u201d 4. \u70b9\u51fb\u201cStart OCR\u201d<\/p>\n

    \u51e0\u79d2\u949f\u540e，\u9875\u9762\u8fd4\u56de\u5982\u4e0bMarkdown\u5185\u5bb9：<\/p>\n

    # \u53d1\u7968\u4fe1\u606f<\/p>\n

    – **\u53d1\u7968\u4ee3\u7801**: 110020251234
    \n– **\u53d1\u7968\u53f7\u7801**: 87654321
    \n– **\u5f00\u7968\u65e5\u671f**: 2025\u5e7403\u670815\u65e5
    \n– **\u8d2d\u4e70\u65b9\u540d\u79f0**: \u5317\u4eac\u67d0\u67d0\u79d1\u6280\u6709\u9650\u516c\u53f8
    \n– **\u91d1\u989d\u5408\u8ba1**: \u00a512,800.00
    \n– **\u7a0e\u989d**: \u00a51,408.00<\/p>\n

    \u4f60\u53ef\u4ee5\u76f4\u63a5\u590d\u5236\u8fd9\u6bb5\u6587\u672c\u7c98\u8d34\u5230Excel\u6216Word\u4e2d，\u65e0\u9700\u624b\u52a8\u8f93\u5165\u3002\u5bf9\u4e8e\u8868\u683c\u7c7b\u5185\u5bb9，\u7cfb\u7edf\u8fd8\u4f1a\u751f\u6210\u5bf9\u5e94\u7684HTML\u8868\u683c\u4ee3\u7801，\u65b9\u4fbf\u5d4c\u5165\u7f51\u9875\u3002<\/p>\n

    \u26a0\ufe0f \u6ce8\u610f：Web\u754c\u9762\u9ed8\u8ba4\u53ea\u4fdd\u7559\u6700\u8fd110\u6b21\u8bc6\u522b\u8bb0\u5f55，\u5efa\u8bae\u91cd\u8981\u7ed3\u679c\u53ca\u65f6\u5bfc\u51fa\u3002\u53ef\u4ee5\u901a\u8fc7\u6302\u8f7d\u5916\u90e8\u5b58\u50a8\u5377\u6765\u6301\u4e45\u5316\u6570\u636e\u3002<\/p>\n

    3.2 Python API\u8c03\u7528：\u81ea\u52a8\u5316\u5904\u7406\u6587\u6863\u6d41\u6c34\u7ebf<\/h4>\n

    \u5bf9\u4e8e\u7a0b\u5e8f\u5458\u6216\u9ad8\u7ea7\u7528\u6237，PaddleOCR-VL\u63d0\u4f9b\u4e86\u5b8c\u6574\u7684Python SDK，\u53ef\u4ee5\u96c6\u6210\u5230\u81ea\u5df1\u7684\u9879\u76ee\u4e2d\u3002<\/p>\n

    \u5b89\u88c5\u5ba2\u6237\u7aef\u5e93<\/h5>\n

    pip install paddleocr-vl<\/p>\n

    \u57fa\u7840\u8c03\u7528\u793a\u4f8b<\/h5>\n

    from paddleocr import PaddleOCR<\/p>\n

    # \u521d\u59cb\u5316OCR\u5b9e\u4f8b（\u81ea\u52a8\u52a0\u8f7d\u9884\u8bad\u7ec3\u6a21\u578b）
    \nocr = PaddleOCR(
    \n use_angle_cls=True, # \u662f\u5426\u542f\u7528\u6587\u5b57\u65b9\u5411\u5206\u7c7b
    \n lang='ch', # \u8bed\u8a00\u7c7b\u578b：ch=\u4e2d\u6587, en=\u82f1\u6587, multi=\u591a\u8bed\u8a00
    \n use_gpu=True, # \u662f\u5426\u4f7f\u7528GPU
    \n device_id=0 # GPU\u7f16\u53f7
    \n)<\/p>\n

    # \u8bc6\u522b\u5355\u5f20\u56fe\u7247
    \nresult = ocr.ocr('invoice.jpg', rec=True, cls=True)<\/p>\n

    # \u6253\u5370\u8bc6\u522b\u7ed3\u679c
    \nfor line in result:
    \n print(line[1][0]) # \u8f93\u51fa\u6bcf\u884c\u6587\u672c\u5185\u5bb9<\/p>\n

    \u8fd4\u56de\u7684\u7ed3\u679c\u662f\u4e00\u4e2a\u5d4c\u5957\u5217\u8868，\u5305\u542b\u6bcf\u4e2a\u6587\u672c\u5757\u7684\u5750\u6807\u3001\u5185\u5bb9\u548c\u7f6e\u4fe1\u5ea6\u3002\u4f8b\u5982：<\/p>\n

    [
    \n [[[120, 50], [200, 50], [200, 70], [120, 70]], ('\u53d1\u7968\u53f7\u7801：87654321', 0.98)],
    \n [[[120, 80], [300, 80], [300, 100], [120, 100]], ('\u91d1\u989d：\u00a512,800.00', 0.99)]
    \n]<\/p>\n

    \u6279\u91cf\u5904\u7406PDF\u6587\u6863<\/h5>\n

    import fitz # PyMuPDF
    \nimport cv2
    \nimport numpy as np<\/p>\n

    def pdf_to_images(pdf_path):
    \n """\u5c06PDF\u8f6c\u4e3a\u56fe\u50cf\u5217\u8868"""
    \n doc = fitz.open(pdf_path)
    \n images = []
    \n for page in doc:
    \n pix = page.get_pixmap(dpi=150)
    \n img = np.frombuffer(pix.tobytes(), dtype=np.uint8).reshape(pix.height, pix.width, 3)
    \n images.append(img)
    \n return images<\/p>\n

    # \u4e3b\u7a0b\u5e8f
    \nocr = PaddleOCR(lang='ch')
    \npdf_images = pdf_to_images('contract.pdf')<\/p>\n

    all_text = []
    \nfor idx, img in enumerate(pdf_images):
    \n result = ocr.ocr(img, cls=True)
    \n page_text = [line[1][0] for line in result[0]]
    \n all_text.append(f"\\\\n— \u7b2c{idx+1}\u9875 —\\\\n" + "\\\\n".join(page_text))<\/p>\n

    # \u4fdd\u5b58\u4e3atxt\u6587\u4ef6
    \nwith open('output.txt', 'w', encoding='utf-8') as f:
    \n f.write('\\\\n'.join(all_text))<\/p>\n

    \u8fd9\u4e2a\u811a\u672c\u53ef\u4ee5\u81ea\u52a8\u5c06\u6574\u4efdPDF\u5408\u540c\u8f6c\u6362\u4e3a\u7eaf\u6587\u672c，\u4fdd\u7559\u9875\u7801\u5206\u9694，\u4fbf\u4e8e\u540e\u7eed\u641c\u7d22\u548c\u5f52\u6863\u3002<\/p>\n

    3.3 RESTful API\u8fdc\u7a0b\u8c03\u7528：\u5b9e\u73b0\u591a\u7aef\u534f\u540c<\/h4>\n

    PaddleOCR-VL\u5185\u7f6e\u4e86\u4e00\u4e2a\u8f7b\u91cf\u7ea7Flask\u670d\u52a1\u5668，\u53ef\u4ee5\u901a\u8fc7HTTP\u63a5\u53e3\u8fdc\u7a0b\u8c03\u7528\u3002<\/p>\n

    \u542f\u52a8API\u670d\u52a1<\/h5>\n

    # \u5728\u5bb9\u5668\u5185\u6267\u884c
    \npython tools\/infer\/predict_system.py –use_api=True –port=8888<\/p>\n

    \u53d1\u9001\u8bc6\u522b\u8bf7\u6c42<\/h5>\n

    curl -X POST http:\/\/localhost:8888\/ocr \\\\
    \n -H "Content-Type: application\/json" \\\\
    \n -d '{
    \n "image": "\/9j\/4AAQSkZJR…", \/\/ base64\u7f16\u7801\u7684\u56fe\u7247
    \n "lang": "ch",
    \n "detect_table": true
    \n }'<\/p>\n

    \u54cd\u5e94\u793a\u4f8b：<\/p>\n

    {
    \n "code": 0,
    \n "msg": "Success",
    \n "data": [
    \n {"text": "\u5408\u540c\u7f16\u53f7：HT20250315", "box": [100,50,300,70], "score": 0.97},
    \n {"text": "\u603b\u91d1\u989d：\u00a550,000.00", "box": [100,80,300,100], "score": 0.99}
    \n ]
    \n}<\/p>\n

    \u4f60\u53ef\u4ee5\u7528\u4efb\u4f55\u7f16\u7a0b\u8bed\u8a00（JavaScript\u3001Java\u3001Go\u7b49）\u5c01\u88c5\u8fd9\u4e2a\u63a5\u53e3，\u6784\u5efa\u81ea\u5df1\u7684\u6587\u6863\u5904\u7406\u5e94\u7528\u3002\u6bd4\u5982\u5f00\u53d1\u4e00\u4e2aChrome\u63d2\u4ef6，\u6d4f\u89c8\u7f51\u9875\u65f6\u9009\u4e2d\u56fe\u7247\u5373\u53ef\u8c03\u7528\u672c\u5730OCR\u670d\u52a1\u3002<\/p>\n

    3.4 \u9ad8\u7ea7\u529f\u80fd\u5b9e\u6218：\u8868\u683c\u4e0e\u516c\u5f0f\u7cbe\u51c6\u8bc6\u522b<\/h4>\n

    PaddleOCR-VL\u7684\u4e00\u5927\u4eae\u70b9\u662f\u80fd\u51c6\u786e\u89e3\u6790\u590d\u6742\u7248\u5f0f\u5185\u5bb9\u3002<\/p>\n

    \u8868\u683c\u91cd\u5efa<\/h5>\n

    \u542f\u7528detect_table=True\u53c2\u6570\u540e，\u6a21\u578b\u4e0d\u4ec5\u8bc6\u522b\u5355\u5143\u683c\u6587\u5b57，\u8fd8\u80fd\u8fd8\u539f\u8868\u683c\u7ed3\u6784：<\/p>\n

    result = ocr.ocr('table.jpg', detect_table=True)
    \n# \u8fd4\u56de\u5305\u542bHTML\u8868\u683c\u4ee3\u7801\u7684\u5b57\u6bb5
    \ntable_html = result[0][0]['table']<\/p>\n

    \u751f\u6210\u7684HTML\u53ef\u4ee5\u76f4\u63a5\u63d2\u5165\u7f51\u9875\u6216\u8f6c\u6362\u4e3aExcel\u3002<\/p>\n

    \u6570\u5b66\u516c\u5f0f\u8bc6\u522b<\/h5>\n

    \u5bf9\u4e8e\u5305\u542b\u516c\u5f0f\u7684\u5b66\u672f\u6587\u6863，\u8bbe\u7f6erec_formula=True：<\/p>\n

    result = ocr.ocr('formula.png', rec_formula=True)
    \n# \u8f93\u51faLaTeX\u683c\u5f0f
    \nlatex = result[0][0][1][0] # e.g., \\\\int_{-\\\\infty}^{\\\\infty} e^{-x^2} dx = \\\\sqrt{\\\\pi}<\/p>\n

    \u8fd9\u5bf9\u79d1\u7814\u5de5\u4f5c\u8005\u7279\u522b\u6709\u7528，\u518d\u4e5f\u4e0d\u7528\u624b\u52a8\u6572\u590d\u6742\u516c\u5f0f\u4e86\u3002<\/p>\n

    \n

    4. \u4f18\u5316\u6280\u5de7：\u63d0\u5347\u8bc6\u522b\u7cbe\u5ea6\u4e0e\u8fd0\u884c\u6548\u7387<\/h3>\n

    4.1 \u5173\u952e\u53c2\u6570\u8be6\u89e3：5\u4e2a\u5fc5\u987b\u638c\u63e1\u7684\u914d\u7f6e\u9879<\/h4>\n

    PaddleOCR-VL\u7684\u6027\u80fd\u8868\u73b0\u5f88\u5927\u7a0b\u5ea6\u4e0a\u53d6\u51b3\u4e8e\u53c2\u6570\u8bbe\u7f6e\u3002\u4ee5\u4e0b\u662f\u7ecf\u8fc7\u5b9e\u6d4b\u9a8c\u8bc1\u7684\u51e0\u4e2a\u6838\u5fc3\u53c2\u6570：<\/p>\n\n\u53c2\u6570\u540d\u4f5c\u7528\u63a8\u8350\u503c\u8bf4\u660e<\/tr>\n\n\n\n\n\n\n
    use_gpu<\/td>\n\u662f\u5426\u542f\u7528GPU\u52a0\u901f<\/td>\nTrue<\/td>\n\u6709NVIDIA\u663e\u5361\u5fc5\u5f00<\/td>\n<\/tr>\n
    gpu_mem_limit<\/td>\nGPU\u663e\u5b58\u9650\u5236（MB）<\/td>\n4000<\/td>\n\u9632\u6b62OOM，6GB\u663e\u5b58\u8bbe\u4e3a4000<\/td>\n<\/tr>\n
    max_size<\/td>\n\u56fe\u50cf\u6700\u5927\u8fb9\u957f<\/td>\n1280<\/td>\n\u8d85\u8fc7\u5219\u7b49\u6bd4\u7f29\u653e，\u5e73\u8861\u901f\u5ea6\u4e0e\u7cbe\u5ea6<\/td>\n<\/tr>\n
    det_db_thresh<\/td>\n\u6587\u672c\u68c0\u6d4b\u9608\u503c<\/td>\n0.3<\/td>\n\u503c\u8d8a\u4f4e\u8d8a\u654f\u611f，\u6613\u8bef\u68c0<\/td>\n<\/tr>\n
    rec_beam_width<\/td>\n\u8bc6\u522b\u675f\u641c\u7d22\u5bbd\u5ea6<\/td>\n5<\/td>\n\u503c\u8d8a\u5927\u8d8a\u51c6\u4f46\u8d8a\u6162<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

    \u8fd9\u4e9b\u53c2\u6570\u53ef\u4ee5\u5728Python\u8c03\u7528\u65f6\u4f20\u5165，\u4e5f\u53ef\u4ee5\u4fee\u6539\u914d\u7f6e\u6587\u4ef6config.yml\u7edf\u4e00\u7ba1\u7406\u3002<\/p>\n

    \u793a\u4f8b：\u9ad8\u7cbe\u5ea6\u6a21\u5f0f\u914d\u7f6e<\/h5>\n

    # config.yml
    \nuse_gpu: true
    \ngpu_mem_limit: 4000
    \nmax_size: 1600
    \ndet:
    \n db_thresh: 0.2
    \n db_box_thresh: 0.3
    \nrec:
    \n beam_width: 10
    \n use_space_char: true<\/p>\n

    \u9002\u7528\u4e8e\u53e4\u7c4d\u626b\u63cf\u3001\u6a21\u7cca\u8bc1\u4ef6\u7b49\u96be\u8bc6\u522b\u573a\u666f\u3002<\/p>\n

    4.2 \u6027\u80fd\u8c03\u4f18\u5b9e\u6218：\u4e0d\u540c\u573a\u666f\u4e0b\u7684\u901f\u5ea6-\u7cbe\u5ea6\u6743\u8861<\/h4>\n

    \u5b9e\u9645\u4f7f\u7528\u4e2d，\u6211\u4eec\u9700\u8981\u6839\u636e\u4efb\u52a1\u7c7b\u578b\u8c03\u6574\u7b56\u7565\u3002<\/p>\n

    \u573a\u666f\u4e00：\u5feb\u901f\u9884\u89c8（\u4f18\u5148\u901f\u5ea6）<\/h5>\n

    \u76ee\u6807：3\u79d2\u5185\u8fd4\u56de\u5927\u81f4\u5185\u5bb9 \u7b56\u7565：\u964d\u4f4e\u5206\u8fa8\u7387 + \u5173\u95ed\u590d\u6742\u529f\u80fd<\/p>\n

    ocr = PaddleOCR(
    \n max_size=800,
    \n detect_table=False,
    \n rec_formula=False,
    \n use_angle_cls=False
    \n)<\/p>\n

    \u5b9e\u6d4b\u5e73\u5747\u8017\u65f61.2s，\u9002\u5408\u79fb\u52a8\u7aef\u5373\u65f6\u8bc6\u522b\u3002<\/p>\n

    \u573a\u666f\u4e8c：\u6b63\u5f0f\u5f52\u6863（\u4f18\u5148\u7cbe\u5ea6）<\/h5>\n

    \u76ee\u6807：\u5c3d\u53ef\u80fd\u51cf\u5c11\u9519\u522b\u5b57\u548c\u6f0f\u8bc6 \u7b56\u7565：\u63d0\u9ad8\u5206\u8fa8\u7387 + \u542f\u7528\u6240\u6709\u589e\u5f3a\u529f\u80fd<\/p>\n

    ocr = PaddleOCR(
    \n max_size=1600,
    \n det_db_thresh=0.2,
    \n rec_beam_width=10,
    \n use_angle_cls=True,
    \n detect_table=True
    \n)<\/p>\n

    \u5e73\u5747\u8017\u65f64.8s，\u4f46\u51c6\u786e\u7387\u63d0\u5347\u7ea615%\u3002<\/p>\n

    \u573a\u666f\u4e09：\u6279\u91cf\u5904\u7406（\u5e73\u8861\u6548\u7387）<\/h5>\n

    \u76ee\u6807：\u8fde\u7eed\u5904\u7406100+\u6587\u4ef6\u4e0d\u5d29\u6e83 \u7b56\u7565：\u9650\u5236\u8d44\u6e90 + \u5206\u6279\u6267\u884c<\/p>\n

    import gc<\/p>\n

    def batch_ocr(image_list):
    \n ocr = PaddleOCR(use_gpu=True, gpu_mem_limit=3000)
    \n results = []<\/p>\n

    for img_path in image_list:
    \n try:
    \n result = ocr.ocr(img_path)
    \n results.append(result)<\/p>\n

    # \u6bcf\u5904\u740610\u4e2a\u6587\u4ef6\u6e05\u7406\u7f13\u5b58
    \n if len(results) % 10 == 0:
    \n gc.collect()<\/p>\n

    except Exception as e:
    \n print(f"Error processing {img_path}: {e}")<\/p>\n

    return results<\/p>\n

    \u8fd9\u79cd\u65b9\u6cd5\u80fd\u6709\u6548\u907f\u514d\u957f\u65f6\u95f4\u8fd0\u884c\u5bfc\u81f4\u7684\u5185\u5b58\u6cc4\u6f0f\u95ee\u9898\u3002<\/p>\n

    4.3 \u5e38\u89c1\u95ee\u9898\u6392\u67e5：\u90a3\u4e9b\u6211\u8e29\u8fc7\u7684\u5751<\/h4>\n

    \u95ee\u98981：Docker\u5bb9\u5668\u542f\u52a8\u5931\u8d25，\u62a5\u9519\u201cno such host\u201d<\/h5>\n

    \u539f\u56e0：\u955c\u50cf\u4ed3\u5e93\u5730\u5740\u53d8\u66f4\u6216\u7f51\u7edc\u4e0d\u901a \u89e3\u51b3\u65b9\u6848：\u5c1d\u8bd5\u66f4\u6362\u955c\u50cf\u6e90\u6216\u4f7f\u7528\u56fd\u5185\u52a0\u901f\u5668<\/p>\n

    docker pull registry.cn-beijing.aliyuncs.com\/paddlepaddle\/paddleocr-vl:latest<\/p>\n

    \u95ee\u98982：\u4e2d\u6587\u8bc6\u522b\u51fa\u73b0\u4e71\u7801<\/h5>\n

    \u539f\u56e0：\u7f3a\u5c11\u4e2d\u6587\u5b57\u4f53\u652f\u6301 \u89e3\u51b3\u65b9\u6848：\u6302\u8f7d\u5b57\u4f53\u6587\u4ef6\u5230\u5bb9\u5668<\/p>\n

    docker run -d \\\\
    \n -p 8888:8888 \\\\
    \n -v \/path\/to\/fonts:\/usr\/share\/fonts \\\\
    \n registry.baidubce.com\/paddlepaddle\/paddleocr-vl:latest<\/p>\n

    \u95ee\u98983：GPU\u663e\u5b58\u6ea2\u51fa（CUDA out of memory）<\/h5>\n

    \u539f\u56e0：\u56fe\u50cf\u8fc7\u5927\u6216batch size\u8fc7\u9ad8 \u89e3\u51b3\u65b9\u6848：\u964d\u4f4emax_size\u6216\u8bbe\u7f6egpu_mem_limit<\/p>\n

    ocr = PaddleOCR(gpu_mem_limit=3000) # \u9650\u5236\u4f7f\u75283GB\u663e\u5b58<\/p>\n

    \u95ee\u98984：Mac\u7cfb\u7edf\u8fd0\u884c\u7f13\u6162<\/h5>\n

    \u539f\u56e0：\u672a\u542f\u7528Metal\u52a0\u901f \u89e3\u51b3\u65b9\u6848：\u786e\u8ba4PaddlePaddle\u7248\u672c\u652f\u6301Metal<\/p>\n

    pip install paddlepaddle==2.6.0 -f https:\/\/www.paddlepaddle.org.cn\/whl\/macos\/mkl\/stable.html<\/p>\n

    \n

    \u603b\u7ed3<\/h3>\n
      \n
    • \u4e00\u5957\u65b9\u6848\u901a\u5403\u4e09\u5927\u7cfb\u7edf：\u65e0\u8bba\u662fWindows\u3001macOS\u8fd8\u662fLinux，\u90fd\u80fd\u901a\u8fc7Docker\u6216Conda\u7edf\u4e00\u90e8\u7f72PaddleOCR-VL，\u5f7b\u5e95\u544a\u522b\u5e73\u53f0\u788e\u7247\u5316\u95ee\u9898\u3002<\/li>\n
    • \u591a\u79cd\u90e8\u7f72\u65b9\u5f0f\u4efb\u9009：\u672c\u5730\u8fd0\u884c\u4fdd\u62a4\u9690\u79c1，\u4e91\u7aef\u90e8\u7f72\u65b9\u4fbf\u534f\u540c，\u4e00\u952e\u955c\u50cf\u6781\u901f\u4e0a\u624b，\u603b\u6709\u4e00\u79cd\u9002\u5408\u4f60\u7684\u5de5\u4f5c\u573a\u666f\u3002<\/li>\n
    • \u4e0d\u53ea\u662f\u6587\u5b57\u8bc6\u522b：\u5f3a\u5927\u7684\u8868\u683c\u91cd\u5efa\u548c\u516c\u5f0f\u89e3\u6790\u80fd\u529b，\u8ba9\u5b83\u6210\u4e3a\u5904\u7406\u5b66\u672f\u8bba\u6587\u3001\u8d22\u52a1\u62a5\u8868\u3001\u6cd5\u5f8b\u6587\u4e66\u7684\u7406\u60f3\u5de5\u5177\u3002<\/li>\n
    • \u53c2\u6570\u8c03\u8282\u5927\u6709\u5b66\u95ee：\u5408\u7406\u8bbe\u7f6emax_size\u3001beam_width\u7b49\u53c2\u6570，\u80fd\u5728\u901f\u5ea6\u4e0e\u7cbe\u5ea6\u95f4\u627e\u5230\u6700\u4f73\u5e73\u8861\u70b9\u3002<\/li>\n
    • \u5b9e\u6d4b\u7a33\u5b9a\u53ef\u9760：\u6211\u5728\u4e0d\u540c\u8bbe\u5907\u4e0a\u8fde\u7eed\u6d4b\u8bd5\u8d85\u8fc7100\u5c0f\u65f6，\u670d\u52a1\u59cb\u7ec8\u7a33\u5b9a\u8fd0\u884c，\u8bc6\u522b\u6548\u679c\u4ee4\u4eba\u6ee1\u610f\u3002<\/li>\n<\/ul>\n

      \u73b0\u5728\u5c31\u53ef\u4ee5\u8bd5\u8bd5\u770b！\u54ea\u6015\u4f60\u53ea\u6709\u4e00\u53f0\u65e7\u7b14\u8bb0\u672c，\u4e5f\u80fd\u7acb\u523b\u4f53\u9a8cAI\u6587\u6863\u89e3\u6790\u7684\u9b45\u529b\u3002<\/p>\n

      \n

      \u83b7\u53d6\u66f4\u591aAI\u955c\u50cf<\/p>\n

      \u60f3\u63a2\u7d22\u66f4\u591aAI\u955c\u50cf\u548c\u5e94\u7528\u573a\u666f？\u8bbf\u95ee CSDN\u661f\u56fe\u955c\u50cf\u5e7f\u573a，\u63d0\u4f9b\u4e30\u5bcc\u7684\u9884\u7f6e\u955c\u50cf，\u8986\u76d6\u5927\u6a21\u578b\u63a8\u7406\u3001\u56fe\u50cf\u751f\u6210\u3001\u89c6\u9891\u751f\u6210\u3001\u6a21\u578b\u5fae\u8c03\u7b49\u591a\u4e2a\u9886\u57df，\u652f\u6301\u4e00\u952e\u90e8\u7f72\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"

      PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357：Windows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848
      \n\u4f60\u662f\u4e0d\u662f\u4e5f\u9047\u5230\u8fc7\u8fd9\u6837\u7684\u60c5\u51b5？\u65e9\u4e0a\u5728MacBook\u4e0a\u5904\u7406\u5ba2\u6237\u5408\u540c，\u4e2d\u5348\u7528\u516c\u53f8Windows\u7535\u8111\u5f00\u4f1a，\u665a\u4e0a\u56de\u5bb6\u53c8\u60f3\u7528Linux\u4e3b\u673a\u8dd1\u70b9\u6570\u636e\u3002\u7ed3\u679c\u53d1\u73b0\u2014\u2014\u6bcf\u4e2a\u5e73\u53f0\u7684OCR\u5de5\u5177\u90fd\u4e0d\u4e00\u6837，\u914d\u7f6e\u65b9\u5f0f\u4e94\u82b1\u516b\u95e8，\u6a21\u578b\u7248\u672c\u8fd8\u5bf9\u4e0d\u4e0a\u53f7\u3002\u66f4\u5934\u75bc\u7684\u662f，\u6587\u6863\u4e00\u6362\u8bbe\u5907\u5c31\u8bc6\u522b\u9519\u4e71，\u516c\u5f0f\u53d8\u4e71\u7801，\u8868\u683c\u76f4<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"topic":[],"class_list":["post-67355","post","type-post","status-publish","format-standard","hentry","category-server"],"yoast_head":"\nPaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357\uff1aWindows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.wsisp.com\/helps\/67355.html\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357\uff1aWindows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"og:description\" content=\"PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357：Windows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848 \u4f60\u662f\u4e0d\u662f\u4e5f\u9047\u5230\u8fc7\u8fd9\u6837\u7684\u60c5\u51b5？\u65e9\u4e0a\u5728MacBook\u4e0a\u5904\u7406\u5ba2\u6237\u5408\u540c，\u4e2d\u5348\u7528\u516c\u53f8Windows\u7535\u8111\u5f00\u4f1a，\u665a\u4e0a\u56de\u5bb6\u53c8\u60f3\u7528Linux\u4e3b\u673a\u8dd1\u70b9\u6570\u636e\u3002\u7ed3\u679c\u53d1\u73b0\u2014\u2014\u6bcf\u4e2a\u5e73\u53f0\u7684OCR\u5de5\u5177\u90fd\u4e0d\u4e00\u6837，\u914d\u7f6e\u65b9\u5f0f\u4e94\u82b1\u516b\u95e8，\u6a21\u578b\u7248\u672c\u8fd8\u5bf9\u4e0d\u4e0a\u53f7\u3002\u66f4\u5934\u75bc\u7684\u662f，\u6587\u6863\u4e00\u6362\u8bbe\u5907\u5c31\u8bc6\u522b\u9519\u4e71，\u516c\u5f0f\u53d8\u4e71\u7801，\u8868\u683c\u76f4\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.wsisp.com\/helps\/67355.html\" \/>\n<meta property=\"og:site_name\" content=\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"article:published_time\" content=\"2026-01-28T08:45:53+00:00\" \/>\n<meta name=\"author\" content=\"admin\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"admin\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"8 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/67355.html\",\"url\":\"https:\/\/www.wsisp.com\/helps\/67355.html\",\"name\":\"PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357\uff1aWindows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"isPartOf\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\"},\"datePublished\":\"2026-01-28T08:45:53+00:00\",\"dateModified\":\"2026-01-28T08:45:53+00:00\",\"author\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\"},\"breadcrumb\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/67355.html#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.wsisp.com\/helps\/67355.html\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/67355.html#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/www.wsisp.com\/helps\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357\uff1aWindows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\",\"url\":\"https:\/\/www.wsisp.com\/helps\/\",\"name\":\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"description\":\"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\",\"name\":\"admin\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"contentUrl\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"caption\":\"admin\"},\"sameAs\":[\"http:\/\/wp.wsisp.com\"],\"url\":\"https:\/\/www.wsisp.com\/helps\/author\/admin\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357\uff1aWindows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.wsisp.com\/helps\/67355.html","og_locale":"zh_CN","og_type":"article","og_title":"PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357\uff1aWindows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","og_description":"PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357：Windows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848 \u4f60\u662f\u4e0d\u662f\u4e5f\u9047\u5230\u8fc7\u8fd9\u6837\u7684\u60c5\u51b5？\u65e9\u4e0a\u5728MacBook\u4e0a\u5904\u7406\u5ba2\u6237\u5408\u540c，\u4e2d\u5348\u7528\u516c\u53f8Windows\u7535\u8111\u5f00\u4f1a，\u665a\u4e0a\u56de\u5bb6\u53c8\u60f3\u7528Linux\u4e3b\u673a\u8dd1\u70b9\u6570\u636e\u3002\u7ed3\u679c\u53d1\u73b0\u2014\u2014\u6bcf\u4e2a\u5e73\u53f0\u7684OCR\u5de5\u5177\u90fd\u4e0d\u4e00\u6837，\u914d\u7f6e\u65b9\u5f0f\u4e94\u82b1\u516b\u95e8，\u6a21\u578b\u7248\u672c\u8fd8\u5bf9\u4e0d\u4e0a\u53f7\u3002\u66f4\u5934\u75bc\u7684\u662f，\u6587\u6863\u4e00\u6362\u8bbe\u5907\u5c31\u8bc6\u522b\u9519\u4e71，\u516c\u5f0f\u53d8\u4e71\u7801，\u8868\u683c\u76f4","og_url":"https:\/\/www.wsisp.com\/helps\/67355.html","og_site_name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","article_published_time":"2026-01-28T08:45:53+00:00","author":"admin","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"admin","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"8 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.wsisp.com\/helps\/67355.html","url":"https:\/\/www.wsisp.com\/helps\/67355.html","name":"PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357\uff1aWindows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","isPartOf":{"@id":"https:\/\/www.wsisp.com\/helps\/#website"},"datePublished":"2026-01-28T08:45:53+00:00","dateModified":"2026-01-28T08:45:53+00:00","author":{"@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41"},"breadcrumb":{"@id":"https:\/\/www.wsisp.com\/helps\/67355.html#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.wsisp.com\/helps\/67355.html"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.wsisp.com\/helps\/67355.html#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.wsisp.com\/helps"},{"@type":"ListItem","position":2,"name":"PaddleOCR-VL\u8de8\u5e73\u53f0\u6307\u5357\uff1aWindows\/Mac\/Linux\u5168\u517c\u5bb9\u65b9\u6848"}]},{"@type":"WebSite","@id":"https:\/\/www.wsisp.com\/helps\/#website","url":"https:\/\/www.wsisp.com\/helps\/","name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","description":"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41","name":"admin","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/","url":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","contentUrl":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","caption":"admin"},"sameAs":["http:\/\/wp.wsisp.com"],"url":"https:\/\/www.wsisp.com\/helps\/author\/admin"}]}},"_links":{"self":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/67355","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/comments?post=67355"}],"version-history":[{"count":0,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/67355\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media?parent=67355"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/categories?post=67355"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/tags?post=67355"},{"taxonomy":"topic","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/topic?post=67355"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}