AI OCR 与文档处理指南：2026年用 GPT-4o Vision / Claude / EasyOCR 打造智能文档处理系统

OCR（光学字符识别）是企业自动化的重要一环。传统 OCR 对复杂文档效果差，2026年用 GPT-4o Vision 和 Claude 3.5 的视觉理解能力，可以识别任何文档、表格、票据。本文详解 AI OCR 的技术方案与代码实现。

OCR 技术方案对比

GPT-4o Vision OCR 实现

import base64
import anthropic

client = anthropic.Anthropic(
    api_key="sk-holysheep-xxx",
    base_url="https://api.holysheep.ai/v1"
)

def ocr_with_gpt4o(image_path: str) -> str:
    """用 GPT-4o Vision 识别文档"""
    with open(image_path, "rb") as f:
        img_data = base64.b64encode(f.read()).decode()

    response = client.messages.create(
        model="gpt-4o",
        max_tokens=4096,
        messages=[{
            "role": "user",
            "content": [
                {"type": "text", "text": "请识别图片中的所有文字，保持原有格式。如果是表格，请用 Markdown 表格格式输出。"},
                {"type": "image", "source": {"type": "base64", "media_type": "image/png", "data": img_data}}
            ]
        }]
    )
    return response.content

# 识别文档
result = ocr_with_gpt4o("invoice.png")
print(result)

Claude 3.5 Vision OCR

def ocr_with_claude(image_path: str) -> dict:
    """用 Claude 3.5 识别文档并提取结构化数据"""
    with open(image_path, "rb") as f:
        img_data = base64.b64encode(f.read()).decode()

    response = client.messages.create(
        model="claude-3-5-sonnet-20241022",
        max_tokens=4096,
        messages=[{
            "role": "user",
            "content": [
                {"type": "text", "text": """请识别图片中的文档内容，并提取以下结构化信息：
1. 文档类型（发票/合同/身份证/营业执照等）
2. 所有文字内容（保持格式）
3. 关键字段（根据文档类型提取对应信息）"""},
                {"type": "image", "source": {"type": "base64", "media_type": "image/png", "data": img_data}}
            ]
        }]
    )
    return response.content

result = ocr_with_claude("business_license.png")
print(result)

本地 OCR 方案（免费）

# pip install easyocr paddlepaddle paddleocr

import easyocr

def local_ocr(image_path: str) -> str:
    """EasyOCR 本地识别（免费，中英文）"""
    reader = easyocr.Reader(['ch_sim', 'en'])
    results = reader.readtext(image_path)

    text = ""
    for bbox, text_line, confidence in results:
        if confidence > 0.5:
            text += text_line + "\n"
    return text

# 使用
result = local_ocr("document.jpg")
print(result)

表格识别与提取

def extract_table_from_image(image_path: str) -> str:
    """识别图片中的表格并输出 Markdown 格式"""
    with open(image_path, "rb") as f:
        img_data = base64.b64encode(f.read()).decode()

    response = client.messages.create(
        model="gpt-4o",
        max_tokens=4096,
        messages=[{
            "role": "user",
            "content": [
                {"type": "text", "text": "请识别图片中的表格，并输出为 Markdown 格式的表格。如果有合并单元格，请在备注中说明。"},
                {"type": "image", "source": {"type": "base64", "media_type": "image/png", "data": img_data}}
            ]
        }]
    )
    return response.content

# 输出 Markdown 表格，可直接导入 Excel
table_md = extract_table_from_image("financial_report.png")
print(table_md)

批量文档处理工作流

import os
from pathlib import Path

def batch_process_documents(folder_path: str, output_folder: str):
    """批量处理文件夹中的所有图片文档"""
    Path(output_folder).mkdir(parents=True, exist_ok=True)

    supported_formats = [".png", ".jpg", ".jpeg", ".pdf"]

    for file_path in Path(folder_path).glob("*"):
        if file_path.suffix.lower() not in supported_formats:
            continue

        print(f"处理：{file_path.name}")

        try:
            # OCR 识别
            text = ocr_with_gpt4o(str(file_path))

            # 保存结果
            output_file = Path(output_folder) / f"{file_path.stem}.txt"
            output_file.write_text(text, encoding="utf-8")

            print(f"  ✅ 已保存：{output_file.name}")

        except Exception as e:
            print(f"  ❌ 失败：{e}")

# 使用
batch_process_documents("./invoices", "./output")

常见文档类型处理方案