{"id":18004,"date":"2025-01-10T22:55:51","date_gmt":"2025-01-10T14:55:51","guid":{"rendered":"https:\/\/www.aisharenet.com\/?p=18004"},"modified":"2025-08-25T00:42:51","modified_gmt":"2025-08-24T16:42:51","slug":"ollama-ocr","status":"publish","type":"post","link":"https:\/\/www.kdjingpai.com\/ja\/ollama-ocr\/","title":{"rendered":"Ollama OCR\uff1a\u4f7f\u7528Ollama\u4e2d\u89c6\u89c9\u6a21\u578b\u63d0\u53d6\u56fe\u50cf\u4e2d\u7684\u6587\u672c"},"content":{"rendered":"<p><a href=\"https:\/\/www.kdjingpai.com\/ja\/ollama\/\">Ollama<\/a> OCR\u662f\u4e00\u4e2a\u5f3a\u5927\u7684\u5149\u5b66\u5b57\u7b26\u8bc6\u522b(OCR)\u5de5\u5177\u5305\uff0c\u5b83\u5229\u7528Ollama\u5e73\u53f0\u63d0\u4f9b\u7684\u6700\u5148\u8fdb\u89c6\u89c9\u8bed\u8a00\u6a21\u578b\u6765\u4ece\u56fe\u50cf\u4e2d\u63d0\u53d6\u6587\u672c\u3002\u8be5\u9879\u76ee\u65e2\u53ef\u4f5c\u4e3aPython\u5305\u4f7f\u7528\uff0c\u4e5f\u63d0\u4f9b\u4e86\u7528\u6237\u53cb\u597d\u7684Streamlit\u7f51\u9875\u5e94\u7528\u7a0b\u5e8f\u754c\u9762\u3002\u5b83\u652f\u6301\u591a\u79cd\u89c6\u89c9\u6a21\u578b\uff0c\u5305\u62ec\u7528\u4e8e\u5b9e\u65f6\u5904\u7406\u7684LLaVA 7B\u548c\u7528\u4e8e\u590d\u6742\u6587\u6863\u7684\u9ad8\u7cbe\u5ea6Llama 3.2 Vision\u6a21\u578b\u3002Ollama OCR\u7684\u7a81\u51fa\u7279\u70b9\u662f\u652f\u6301\u591a\u79cd\u8f93\u51fa\u683c\u5f0f\uff0c\u5305\u62ecMarkdown\u3001\u7eaf\u6587\u672c\u3001JSON\u7b49\uff0c\u5e76\u4e14\u5177\u5907\u6279\u91cf\u5904\u7406\u80fd\u529b\u3002\u8be5\u5de5\u5177\u7279\u522b\u9002\u5408\u9700\u8981\u4ece\u56fe\u50cf\u4e2d\u63d0\u53d6\u548c\u7ed3\u6784\u5316\u6587\u672c\u6570\u636e\u7684\u5f00\u53d1\u8005\u548c\u7814\u7a76\u4eba\u5458\u4f7f\u7528\u3002<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-18005\" title=\"Ollama OCR\uff1a\u4f7f\u7528Ollama\u4e2d\u89c6\u89c9\u6a21\u578b\u63d0\u53d6\u56fe\u50cf\u4e2d\u7684\u6587\u672c-1\" src=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/01\/e0d45f090a37ede.jpg\" alt=\"Ollama OCR\uff1a\u4f7f\u7528Ollama\u4e2d\u89c6\u89c9\u6a21\u578b\u63d0\u53d6\u56fe\u50cf\u4e2d\u7684\u6587\u672c-1\" width=\"1897\" height=\"933\" srcset=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/01\/e0d45f090a37ede.jpg 1897w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/01\/e0d45f090a37ede-300x148.jpg 300w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/01\/e0d45f090a37ede-1024x504.jpg 1024w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/01\/e0d45f090a37ede-768x378.jpg 768w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/01\/e0d45f090a37ede-1536x755.jpg 1536w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/01\/e0d45f090a37ede-18x9.jpg 18w\" sizes=\"auto, (max-width: 1897px) 100vw, 1897px\" \/><\/p>\n<p>&nbsp;<\/p>\n<h2>\u529f\u80fd\u5217\u8868<\/h2>\n<ul>\n<li>\u652f\u6301\u591a\u79cd\u5148\u8fdb\u89c6\u89c9\u8bed\u8a00\u6a21\u578b\uff08LLaVA 7B\u548cLlama 3.2 Vision\uff09<\/li>\n<li>\u63d0\u4f9b\u591a\u6837\u5316\u7684\u8f93\u51fa\u683c\u5f0f\uff08Markdown\u3001\u7eaf\u6587\u672c\u3001JSON\u3001\u7ed3\u6784\u5316\u6570\u636e\u3001\u952e\u503c\u5bf9\uff09<\/li>\n<li>\u652f\u6301\u6279\u91cf\u56fe\u50cf\u5904\u7406\u529f\u80fd\uff0c\u53ef\u5e76\u884c\u5904\u7406\u591a\u4e2a\u56fe\u50cf<\/li>\n<li>\u5185\u7f6e\u56fe\u50cf\u9884\u5904\u7406\u529f\u80fd\uff08\u8c03\u6574\u5927\u5c0f\u3001\u6807\u51c6\u5316\u7b49\uff09<\/li>\n<li>\u63d0\u4f9b\u8fdb\u5ea6\u8ddf\u8e2a\u548c\u5904\u7406\u7edf\u8ba1\u529f\u80fd<\/li>\n<li>\u652f\u6301\u7528\u6237\u53cb\u597d\u7684Streamlit\u7f51\u9875\u754c\u9762<\/li>\n<li>\u652f\u6301\u62d6\u653e\u5f0f\u56fe\u50cf\u4e0a\u4f20\u548c\u5b9e\u65f6\u5904\u7406<\/li>\n<li>\u63d0\u4f9b\u63d0\u53d6\u6587\u672c\u7684\u4e0b\u8f7d\u529f\u80fd<\/li>\n<li>\u96c6\u6210\u56fe\u50cf\u9884\u89c8\u548c\u8be6\u7ec6\u4fe1\u606f\u663e\u793a<\/li>\n<\/ul>\n<p>&nbsp;<\/p>\n<h2>\u4f7f\u7528\u5e2e\u52a9<\/h2>\n<h3>1. \u5b89\u88c5\u6b65\u9aa4<\/h3>\n<ol>\n<li>\u9996\u5148\u9700\u8981\u5b89\u88c5Ollama\u5e73\u53f0\uff1a\n<ul>\n<li>\u8bbf\u95eeOllama\u5b98\u65b9\u7f51\u7ad9\u4e0b\u8f7d\u5bf9\u5e94\u7cfb\u7edf\u7684\u5b89\u88c5\u5305<\/li>\n<li>\u5b8c\u6210Ollama\u7684\u57fa\u7840\u5b89\u88c5<\/li>\n<\/ul>\n<\/li>\n<li>\u5b89\u88c5\u6240\u9700\u7684\u89c6\u89c9\u6a21\u578b\uff1a<\/li>\n<\/ol>\n<pre><code>ollama pull llama3.2-vision:11b\r\n<\/code><\/pre>\n<ol start=\"3\">\n<li>\u5b89\u88c5Ollama OCR\u5305\uff1a<\/li>\n<\/ol>\n<pre><code>pip install ollama-ocr\r\n<\/code><\/pre>\n<h3>2. Python\u5305\u4f7f\u7528\u65b9\u6cd5<\/h3>\n<h4>2.1 \u5355\u56fe\u50cf\u5904\u7406<\/h4>\n<pre><code>from ollama_ocr import OCRProcessor\r\n# \u521d\u59cb\u5316OCR\u5904\u7406\u5668\r\nocr = OCRProcessor(model_name='llama3.2-vision:11b')\r\n# \u5904\u7406\u5355\u5f20\u56fe\u50cf\r\nresult = ocr.process_image(\r\nimage_path=\"\u56fe\u7247\u8def\u5f84.png\",\r\nformat_type=\"markdown\"  # \u53ef\u9009\u683c\u5f0f\uff1amarkdown, text, json, structured, key_value\r\n)\r\nprint(result)\r\n<\/code><\/pre>\n<h4>2.2 \u6279\u91cf\u5904\u7406\u56fe\u50cf<\/h4>\n<pre><code># \u521d\u59cb\u5316OCR\u5904\u7406\u5668\uff0c\u8bbe\u7f6e\u5e76\u884c\u5904\u7406\u6570\r\nocr = OCRProcessor(model_name='llama3.2-vision:11b', max_workers=4)\r\n# \u6279\u91cf\u5904\u7406\u56fe\u50cf\r\nbatch_results = ocr.process_batch(\r\ninput_path=\"\u56fe\u7247\u6587\u4ef6\u5939\u8def\u5f84\",\r\nformat_type=\"markdown\",\r\nrecursive=True,  # \u641c\u7d22\u5b50\u76ee\u5f55\r\npreprocess=True  # \u542f\u7528\u56fe\u50cf\u9884\u5904\u7406\r\n)\r\n# \u67e5\u770b\u5904\u7406\u7ed3\u679c\r\nfor file_path, text in batch_results['results'].items():\r\nprint(f\"\\n\u6587\u4ef6: {file_path}\")\r\nprint(f\"\u63d0\u53d6\u7684\u6587\u672c: {text}\")\r\n# \u67e5\u770b\u5904\u7406\u7edf\u8ba1\r\nprint(f\"\u603b\u56fe\u50cf\u6570: {batch_results['statistics']['total']}\")\r\nprint(f\"\u6210\u529f\u5904\u7406: {batch_results['statistics']['successful']}\")\r\nprint(f\"\u5904\u7406\u5931\u8d25: {batch_results['statistics']['failed']}\")\r\n<\/code><\/pre>\n<h3>3. Streamlit\u7f51\u9875\u5e94\u7528\u4f7f\u7528\u65b9\u6cd5<\/h3>\n<ol>\n<li>\u514b\u9686\u4ee3\u7801\u4ed3\u5e93\uff1a<\/li>\n<\/ol>\n<pre><code>git clone https:\/\/github.com\/imanoop7\/Ollama-OCR.git\r\ncd Ollama-OCR\r\n<\/code><\/pre>\n<ol start=\"2\">\n<li>\u5b89\u88c5\u4f9d\u8d56\uff1a<\/li>\n<\/ol>\n<pre><code>pip install -r requirements.txt\r\n<\/code><\/pre>\n<ol start=\"3\">\n<li>\u542f\u52a8\u7f51\u9875\u5e94\u7528\uff1a<\/li>\n<\/ol>\n<pre><code>cd src\/ollama_ocr\r\nstreamlit run app.py\r\n<\/code><\/pre>\n<h3>4. \u8f93\u51fa\u683c\u5f0f\u8bf4\u660e<\/h3>\n<ul>\n<li>Markdown\u683c\u5f0f\uff1a\u4fdd\u7559\u6587\u672c\u683c\u5f0f\uff0c\u5305\u62ec\u6807\u9898\u548c\u5217\u8868<\/li>\n<li>\u7eaf\u6587\u672c\u683c\u5f0f\uff1a\u63d0\u4f9b\u5e72\u51c0\u7b80\u6d01\u7684\u6587\u672c\u63d0\u53d6<\/li>\n<li>JSON\u683c\u5f0f\uff1a\u7ed3\u6784\u5316\u7684\u6570\u636e\u683c\u5f0f\u8f93\u51fa<\/li>\n<li>\u7ed3\u6784\u5316\u683c\u5f0f\uff1a\u8868\u683c\u548c\u7ec4\u7ec7\u5316\u6570\u636e<\/li>\n<li>\u952e\u503c\u5bf9\u683c\u5f0f\uff1a\u63d0\u53d6\u5e26\u6807\u7b7e\u7684\u4fe1\u606f<\/li>\n<\/ul>\n<h3>5. \u6ce8\u610f\u4e8b\u9879<\/h3>\n<ul>\n<li>LLaVA\u6a21\u578b\u53ef\u80fd\u5076\u5c14\u4f1a\u4ea7\u751f\u9519\u8bef\u8f93\u51fa\uff0c\u5efa\u8bae\u91cd\u8981\u573a\u666f\u4f7f\u7528Llama 3.2 Vision\u6a21\u578b<\/li>\n<li>\u56fe\u50cf\u9884\u5904\u7406\u53ef\u4ee5\u63d0\u9ad8\u8bc6\u522b\u51c6\u786e\u7387<\/li>\n<li>\u6279\u91cf\u5904\u7406\u65f6\u6ce8\u610f\u5408\u7406\u8bbe\u7f6e\u5e76\u884c\u6570\uff0c\u907f\u514d\u5185\u5b58\u5360\u7528\u8fc7\u9ad8<\/li>\n<li>\u5904\u7406\u5927\u91cf\u56fe\u50cf\u65f6\u5efa\u8bae\u5f00\u542f\u8fdb\u5ea6\u8ddf\u8e2a\u529f\u80fd<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>Ollama OCR\u662f\u4e00\u4e2a\u5f3a\u5927\u7684\u5149\u5b66\u5b57\u7b26\u8bc6\u522b(OCR)\u5de5\u5177\u5305\uff0c\u5b83\u5229\u7528Ollama\u5e73\u53f0\u63d0\u4f9b\u7684\u6700\u5148\u8fdb\u89c6\u89c9\u8bed\u8a00\u6a21\u578b\u6765\u4ece\u56fe\u50cf\u4e2d\u63d0\u53d6\u6587\u672c\u3002\u8be5\u9879\u76ee\u65e2\u53ef\u4f5c\u4e3aPython\u5305\u4f7f\u7528\uff0c\u4e5f\u63d0\u4f9b\u4e86\u7528\u6237\u53cb\u597d\u7684Streamlit\u7f51\u9875\u5e94\u7528\u7a0b\u5e8f\u754c\u9762\u3002\u5b83\u652f\u6301\u591a\u79cd\u89c6\u89c9\u6a21\u578b\uff0c\u5305\u62ec&#8230;<\/p>\n","protected":false},"author":1,"featured_media":32782,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20,502,499],"tags":[230,248,252],"class_list":["post-18004","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tool","category-text-recognition","category-document-extraction","tag-aikaiyuanxiangmu","tag-ocr","tag-markdown"],"_links":{"self":[{"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/posts\/18004","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/comments?post=18004"}],"version-history":[{"count":0,"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/posts\/18004\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/media\/32782"}],"wp:attachment":[{"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/media?parent=18004"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/categories?post=18004"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/ja\/wp-json\/wp\/v2\/tags?post=18004"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}