OverviewHistoryStatsSecurity
npx skills add ...
Documentation
SKILL.md
npx skills add yejinlei/pdf-ocr-skill --skill pdf-ocr
支持四引擎的PDF OCR识别技能，可从影印版PDF文件和图片文件中提取中英文文字内容 | PDF OCR Skill with quadruple-engine support, capable of extracting Chinese and English text from scanned PDF files and image files
npx skills add yejinlei/pdf-ocr-skill --skill pdf-ocr
PDF OCR技能用于从影印版PDF文件和图片文件中提取文字内容。该技能支持两种OCR引擎：
安装RapidOCR以获得本地识别能力：
.env.example 文件并重命名为 .envRapidOCR引擎：
硅基流动API引擎：
RapidDoc引擎：
PaddleOCR引擎：
对于复杂的扫描版PDF或图片，识别准确率可能会有所不同
建议使用高清晰度的扫描版PDF或图片以获得更好的识别效果
在与 AI IDE 中的助手交互时，您可以使用以下提示词来指定使用不同的 OCR 引擎：
示例 1：使用本地引擎
示例 2：使用云端引擎
示例 3：自动选择
当 AI 助手接收到这些提示词时，会：
通过使用这些提示词，您可以在与 AI IDE 交互时灵活控制 OCR 引擎的选择，获得最佳的识别效果
RapidOCR初始化失败
ModuleNotFoundError: No module named 'rapidocr_onnxruntime'pip install rapidocr_onnxruntime硅基流动API 401错误
Unauthorized: 401 Client Error.env文件中PDF转图片失败
ImportError: No module named 'fitz'pip install pymupdf识别结果为空
PDF OCR Skill is used to extract text content from scanned PDF files and image files. This skill supports two OCR engines:
Install RapidOCR for local recognition capability:
.env.example file and rename it to .envRapidOCR Engine:
SiliconFlow API Engine:
RapidDoc Engine:
PaddleOCR Engine:
Recognition accuracy may vary for complex scanned PDFs or images
It is recommended to use high-resolution scanned PDFs or images for better recognition results
When interacting with assistants in AI IDEs, you can use the following prompt words to specify different OCR engines:
Example 1: Using Local Engine
Example 2: Using Cloud Engine
Example 3: Automatic Selection
When the AI assistant receives these prompt words, it will:
By using these prompt words, you can flexibly control the OCR engine selection when interacting with AI IDEs to get the best recognition results
RapidOCR Initialization Failure
ModuleNotFoundError: No module named 'rapidocr_onnxruntime'pip install rapidocr_onnxruntimeSiliconFlow API 401 Error
Unauthorized: 401 Client Error.env filePDF to Image Conversion Failure
ImportError: No module named 'fitz'pip install pymupdfEmpty Recognition Result
MIT License - See LICENSE.txt
# 导入OCR处理器
from scripts.pdf_ocr_processor import PDFOCRProcessor
# 创建处理器实例（默认使用RapidOCR）
processor = PDFOCRProcessor()
# 执行PDF OCR识别
result = processor.ocr_pdf('path/to/your/scanned.pdf')
# 获取识别结果
print(f"识别完成，共 {result['page_count']} 页")
print(f"使用引擎: {result['engine']}")
print(result['text'])# 导入OCR处理器
from scripts.pdf_ocr_processor import PDFOCRProcessor
# 创建处理器实例，指定使用硅基流动API
processor = PDFOCRProcessor(engine="siliconflow")
# 执行PDF OCR识别
result = processor.ocr_pdf('path/to/your/scanned.pdf')
# 获取识别结果
print(f"识别完成，共 {result['page_count']} 页")
print(result['text'])# 导入OCR处理器
from scripts.pdf_ocr_processor import PDFOCRProcessor
# 创建处理器实例
processor = PDFOCRProcessor() # 或 PDFOCRProcessor(engine="siliconflow")
# 执行图片OCR识别
result = processor.ocr_image_file('path/to/your/image.jpg')
# 获取识别结果
print(f"识别结果: {result['text']}")# 使用默认RapidOCR引擎
python pdf_ocr_processor.py your_document.pdf
# 使用硅基流动API引擎
python pdf_ocr_processor.py your_document.pdf siliconflow
# 使用RapidDoc增强引擎
python pdf_ocr_processor.py your_document.pdf rapidoc
# 使用PaddleOCR引擎
python pdf_ocr_processor.py your_document.pdf paddleimport os
from scripts.pdf_ocr_processor import PDFOCRProcessor
# 创建处理器实例
processor = PDFOCRProcessor()
# 批量处理目录中的所有PDF文件
pdf_dir = "path/to/pdf/files"
output_dir = "path/to/output"
os.makedirs(output_dir, exist_ok=True)
for pdf_file in os.listdir(pdf_dir):
if pdf_file.endswith('.pdf'):
pdf_path = os.path.join(pdf_dir, pdf_file)
output_path = os.path.join(output_dir, f"{os.path.splitext(pdf_file)[0]}.txt")
print(f"处理文件: {pdf_file}")
try:
result = processor.ocr_pdf(pdf_path)
# 保存识别结果到文本文件
with open(output_path, 'w', encoding='utf-8') as f:
f.write(f"=== PDF OCR 识别结果 ===\n")
f.write(f"文件名: {pdf_file}\n")
f.write(f"页数: {result['page_count']}\n")
f.write(f"使用引擎: {result['engine']}\n\n")
f.write(result['text'])
print(f"处理完成，结果已保存到: {output_path}")
except Exception as e:
print(f"处理失败: {e}")from scripts.pdf_ocr_processor import PDFOCRProcessor
def process_with_best_engine(pdf_path):
"""尝试使用RapidOCR，如果效果不佳则使用硅基流动API"""
# 首先使用RapidOCR本地引擎
rapid_processor = PDFOCRProcessor(engine="rapid")
rapid_result = rapid_processor.ocr_pdf(pdf_path)
# 简单评估识别效果（例如：检查识别出的文本长度）
text_length = len(rapid_result['text'])
if text_length < 100: # 如果识别出的文本太短，可能效果不佳
print("RapidOCR识别效果可能不佳，尝试使用硅基流动API...")
silicon_processor = PDFOCRProcessor(engine="siliconflow")
silicon_result = silicon_processor.ocr_pdf(pdf_path)
return silicon_result
else:
return rapid_result
# 使用示例
result = process_with_best_engine('path/to/your/document.pdf')
print(f"识别完成，使用引擎: {result['engine']}")
print(result['text']){
"text": "识别的完整文本内容",
"page_count": 页数, # 图片文件始终为1
"engine": "rapid" | "rapidoc" | "paddle" | "siliconflow" # 使用的OCR引擎
}用户：帮我处理这个扫描版 PDF，用本地 OCR 引擎快速识别
助手：好的，我将使用 RapidOCR 本地引擎为您处理。请提供 PDF 文件路径。用户：这个 PDF 包含手写体，需要高精度识别，用硅基流动 API
助手：理解，我将使用硅基流动 API 大模型为您处理。请提供 PDF 文件路径和您的 API 密钥（如果尚未配置）。用户：帮我识别这个 PDF，选择最合适的引擎
助手：我将默认使用 RapidOCR 本地引擎为您处理。如果识别效果不理想，我们可以尝试使用硅基流动 API。# OCR engine selection
# - "rapid": Use RapidOCR local engine (default, no API key required)
# - "rapidoc": Use RapidDoc enhanced engine (no API key required)
# - "paddle": Use PaddleOCR local engine (no API key required)
# - "siliconflow": Use SiliconFlow API engine (API key required)
OCR_ENGINE=rapid
# If using SiliconFlow API engine, configure the following options:
SILICON_FLOW_API_KEY=your_api_key_here
SILICON_FLOW_OCR_MODEL=deepseek-ai/DeepSeek-OCR# Import OCR processor
from scripts.pdf_ocr_processor import PDFOCRProcessor
# Create processor instance (default uses RapidOCR)
processor = PDFOCRProcessor()
# Perform PDF OCR recognition
result = processor.ocr_pdf('path/to/your/scanned.pdf')
# Get recognition result
print(f"Recognition completed, total {result['page_count']} pages")
print(f"Engine used: {result['engine']}")
print(result['text'])# Import OCR processor
from scripts.pdf_ocr_processor import PDFOCRProcessor
# Create processor instance, specify to use SiliconFlow API
processor = PDFOCRProcessor(engine="siliconflow")
# Perform PDF OCR recognition
result = processor.ocr_pdf('path/to/your/scanned.pdf')
# Get recognition result
print(f"Recognition completed, total {result['page_count']} pages")
print(result['text'])# Import OCR processor
from scripts.pdf_ocr_processor import PDFOCRProcessor
# Create processor instance
processor = PDFOCRProcessor() # or PDFOCRProcessor(engine="siliconflow")
# Perform image OCR recognition
result = processor.ocr_image_file('path/to/your/image.jpg')
# Get recognition result
print(f"Recognition result: {result['text']}")# Use default RapidOCR engine
python pdf_ocr_processor.py your_document.pdf
# Use SiliconFlow API engine
python pdf_ocr_processor.py your_document.pdf siliconflow
# Use RapidDoc enhanced engine
python pdf_ocr_processor.py your_document.pdf rapidoc
# Use PaddleOCR engine
python pdf_ocr_processor.py your_document.pdf paddleimport os
from scripts.pdf_ocr_processor import PDFOCRProcessor
# Create processor instance
processor = PDFOCRProcessor()
# Batch process all PDF files in directory
pdf_dir = "path/to/pdf/files"
output_dir = "path/to/output"
os.makedirs(output_dir, exist_ok=True)
for pdf_file in os.listdir(pdf_dir):
if pdf_file.endswith('.pdf'):
pdf_path = os.path.join(pdf_dir, pdf_file)
output_path = os.path.join(output_dir, f"{os.path.splitext(pdf_file)[0]}.txt")
print(f"Processing file: {pdf_file}")
try:
result = processor.ocr_pdf(pdf_path)
# Save recognition result to text file
with open(output_path, 'w', encoding='utf-8') as f:
f.write(f"=== PDF OCR Recognition Result ===\n")
f.write(f"File name: {pdf_file}\n")
f.write(f"Pages: {result['page_count']}\n")
f.write(f"Engine used: {result['engine']}\n\n")
f.write(result['text'])
print(f"Processing completed, result saved to: {output_path}")
except Exception as e:
print(f"Processing failed: {e}")from scripts.pdf_ocr_processor import PDFOCRProcessor
def process_with_best_engine(pdf_path):
"""Try using RapidOCR, if not good enough then use SiliconFlow API"""
# First use RapidOCR local engine
rapid_processor = PDFOCRProcessor(engine="rapid")
rapid_result = rapid_processor.ocr_pdf(pdf_path)
# Simple evaluation of recognition effect (e.g., check recognized text length)
text_length = len(rapid_result['text'])
if text_length < 100: # If recognized text is too short, may not be good enough
print("RapidOCR recognition effect may not be good enough, trying SiliconFlow API...")
silicon_processor = PDFOCRProcessor(engine="siliconflow")
silicon_result = silicon_processor.ocr_pdf(pdf_path)
return silicon_result
else:
return rapid_result
# Usage example
result = process_with_best_engine('path/to/your/document.pdf')
print(f"Recognition completed, engine used: {result['engine']}")
print(result['text']){
"text": "Recognized full text content",
"page_count": number_of_pages, # Always 1 for image files
"engine": "rapid" | "rapidoc" | "paddle" | "siliconflow" # OCR engine used
}User: Help me process this scanned PDF, use local OCR engine for quick recognition
Assistant: Sure, I'll use the RapidOCR local engine for you. Please provide the PDF file path.User: This PDF contains handwritten text, need high-precision recognition, use SiliconFlow API
Assistant: Understood, I'll use the SiliconFlow API large model for you. Please provide the PDF file path and your API key (if not already configured).User: Help me recognize this PDF, choose the most suitable engine
Assistant: I'll default to using the RapidOCR local engine for you. If the recognition effect is not ideal, we can try using SiliconFlow API.