OCR MCP Service

#!/usr/bin/env python3 """生成批量OCR处理汇总报告。 从ocr_results目录读取所有OCR结果，生成汇总报告。 """ import sys import json import argparse from pathlib import Path from typing import List, Dict from datetime import datetime # Add project root to path project_root = Path(__file__).parent.parent sys.path.insert(0, str(project_root)) from scripts.common import setup_script def load_ocr_results(results_dir: Path) -> List[Dict]: """加载所有OCR结果。""" results = [] for json_file in sorted(results_dir.glob("*_ocr.json")): try: # 尝试读取JSON文件 result = None try: with open(json_file, "r", encoding="utf-8") as f: result = json.load(f) except json.JSONDecodeError: # JSON解析失败，尝试从txt文件读取基本信息 print(f"⚠️ JSON格式错误: {json_file.name}，尝试从txt文件读取...") txt_file = json_file.with_suffix(".txt") if txt_file.exists(): with open(txt_file, "r", encoding="utf-8") as f: text = f.read() # 创建基本结果 result = { "text": text, "engine": "unknown", "confidence": 0.0, "processing_time": 0.0, "boxes": [] } else: print(f" ⏭️ 跳过: {json_file.name}（无法读取）") continue if result: result["source_file"] = json_file.stem.replace("_ocr", "") # 确保必要字段存在 result.setdefault("text", "") result.setdefault("boxes", []) result.setdefault("confidence", 0.0) result.setdefault("processing_time", 0.0) result.setdefault("engine", "unknown") results.append(result) except Exception as e: print(f"⚠️ 无法读取 {json_file.name}: {e}") return results def generate_summary(results: List[Dict], output_file: Path): """生成汇总报告。""" total_images = len(results) total_text_length = sum(len(r.get("text", "")) for r in results) total_boxes = sum(len(r.get("boxes", [])) for r in results) total_processing_time = sum(r.get("processing_time", 0.0) for r in results) avg_confidence = ( sum(r.get("confidence", 0.0) for r in results) / total_images if total_images > 0 else 0.0 ) # 统计信息 summary = { "生成时间": datetime.now().isoformat(), "统计信息": { "总图片数": total_images, "总文本长度": total_text_length, "总文本块数": total_boxes, "总处理时间": f"{total_processing_time:.2f}秒", "平均置信度": f"{avg_confidence:.3f}", "平均处理时间": f"{total_processing_time / total_images:.2f}秒" if total_images > 0 else "0秒" }, "详细结果": [] } # 添加每张图片的详细信息 for result in results: summary["详细结果"].append({ "图片": result["source_file"], "引擎": result.get("engine", "unknown"), "文本长度": len(result.get("text", "")), "文本块数": len(result.get("boxes", [])), "置信度": result.get("confidence", 0.0), "处理时间": f"{result.get('processing_time', 0.0):.2f}秒" }) # 保存JSON格式 with open(output_file, "w", encoding="utf-8") as f: json.dump(summary, f, ensure_ascii=False, indent=2) # 生成Markdown格式 md_file = output_file.with_suffix(".md") with open(md_file, "w", encoding="utf-8") as f: f.write("# OCR批量处理汇总报告\n\n") f.write(f"**生成时间**: {summary['生成时间']}\n\n") f.write("## 统计信息\n\n") stats = summary["统计信息"] f.write(f"- **总图片数**: {stats['总图片数']}\n") f.write(f"- **总文本长度**: {stats['总文本长度']} 字符\n") f.write(f"- **总文本块数**: {stats['总文本块数']}\n") f.write(f"- **总处理时间**: {stats['总处理时间']}\n") f.write(f"- **平均置信度**: {stats['平均置信度']}\n") f.write(f"- **平均处理时间**: {stats['平均处理时间']}\n\n") f.write("## 详细结果\n\n") f.write("| 图片 | 引擎 | 文本长度 | 文本块数 | 置信度 | 处理时间 |\n") f.write("|------|------|----------|----------|--------|----------|\n") for detail in summary["详细结果"]: f.write(f"| {detail['图片']} | {detail['引擎']} | " f"{detail['文本长度']} | {detail['文本块数']} | " f"{detail['置信度']:.3f} | {detail['处理时间']} |\n") print(f"✅ 汇总报告已生成:") print(f" - JSON: {output_file}") print(f" - Markdown: {md_file}") def main(): """主函数。""" parser = argparse.ArgumentParser( description="生成批量OCR处理汇总报告" ) parser.add_argument( "results_dir", type=str, nargs="?", default="ocr_results", help="OCR结果目录（默认：ocr_results）" ) parser.add_argument( "--output", type=str, help="输出文件路径（默认：results_dir/batch_summary）" ) args = parser.parse_args() # 解析路径 if Path(args.results_dir).is_absolute(): results_dir = Path(args.results_dir) else: # 相对路径，尝试从当前目录和项目根目录查找 current_dir = Path.cwd() project_root = Path(__file__).parent.parent for base_dir in [current_dir, project_root]: candidate = base_dir / args.results_dir if candidate.exists(): results_dir = candidate break else: results_dir = Path(args.results_dir) if not results_dir.exists(): print(f"❌ 错误: 目录不存在: {results_dir}") sys.exit(1) # 确定输出文件 if args.output: output_file = Path(args.output) else: output_file = results_dir / "batch_summary.json" # 加载结果 print(f"📂 读取OCR结果: {results_dir}") results = load_ocr_results(results_dir) if not results: print("❌ 未找到OCR结果文件") sys.exit(1) print(f"✅ 找到 {len(results)} 个结果文件") # 生成汇总 generate_summary(results, output_file) if __name__ == "__main__": main()