check_model.py

#!/usr/bin/env python3
"""
模型检查脚本 - 验证单鼠姿态检测模型是否正确加载
"""

import os
import sys
import logging
import numpy as np
import cv2

# 配置日志
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
logger = logging.getLogger(__name__)

def check_model():
    """检查模型文件"""
    base_model_path = "models/kunin-mice-pose.v0.1.5n.pt"
    
    # 选择模型路径（与SingleMouseProcessor保持一致）
    model_path = select_model_path(base_model_path)
    
    logger.info(f"检查模型文件: {model_path}")
    
    # 1. 检查文件是否存在
    if not os.path.exists(model_path):
        logger.error(f"❌ 模型文件不存在: {model_path}")
        return False
    
    logger.info(f"✅ 模型文件存在: {model_path}")
    
    # 2. 检查文件大小
    file_size = os.path.getsize(model_path)
    logger.info(f"📄 文件大小: {file_size / (1024*1024):.2f} MB")
    
    # 3. 检查依赖
    try:
        import torch
        import ultralytics
        from ultralytics import YOLO
        logger.info("✅ 依赖包检查通过")
        
        # 检查CUDA
        if torch.cuda.is_available():
            logger.info(f"✅ CUDA可用: {torch.cuda.get_device_name(0)}")
        else:
            logger.info("⚠️ CUDA不可用，将使用CPU")
            
    except ImportError as e:
        logger.error(f"❌ 缺少依赖包: {str(e)}")
        return False
    
    # 4. 尝试加载模型
    try:
        logger.info("正在加载YOLO模型...")
        model = YOLO(model_path, task="pose")
        logger.info("✅ 模型加载成功")
    except Exception as e:
        logger.error(f"❌ 模型加载失败: {str(e)}")
        import traceback
        logger.error(traceback.format_exc())
        return False
    
    # 5. 测试模型推理
    try:
        logger.info("测试模型推理...")
        
        # 创建测试图像
        test_img = np.random.randint(0, 255, (640, 640, 3), dtype=np.uint8)
        
        # 进行推理
        results = model(test_img, conf=0.3, verbose=False)
        
        logger.info("✅ 模型推理测试成功")
        
        # 检查输出结构
        result = results[0]
        logger.info(f"📊 模型输出信息:")
        logger.info(f"  - 是否有boxes属性: {hasattr(result, 'boxes')}")
        logger.info(f"  - 是否有keypoints属性: {hasattr(result, 'keypoints')}")
        
        if hasattr(result, 'keypoints') and result.keypoints is not None:
            logger.info(f"  - Keypoints shape: {result.keypoints.data.shape}")
        
        return True
        
    except Exception as e:
        logger.error(f"❌ 模型推理测试失败: {str(e)}")
        import traceback
        logger.error(traceback.format_exc())
        return False

def select_model_path(base_model_path: str) -> str:
    """根据GPU情况选择模型路径"""
    try:
        import torch
        # 检测是否有NVIDIA GPU
        if torch.cuda.is_available():
            nvidia_gpu_found = False
            for i in range(torch.cuda.device_count()):
                gpu_name = torch.cuda.get_device_name(i).lower()
                if 'nvidia' in gpu_name:
                    nvidia_gpu_found = True
                    break
            
            if nvidia_gpu_found:
                # 构建.engine模型路径
                engine_path = base_model_path.replace('.pt', '.engine')
                if os.path.exists(engine_path):
                    logger.info(f"🚀 检测到NVIDIA GPU，使用TensorRT模型: {engine_path}")
                    return engine_path
                else:
                    logger.info(f"⚠️ NVIDIA GPU已检测到，但TensorRT模型不存在: {engine_path}")
                    logger.info(f"📍 使用PyTorch模型: {base_model_path}")
                    return base_model_path
            else:
                logger.info(f"📍 检测到GPU但非NVIDIA，使用PyTorch模型: {base_model_path}")
                return base_model_path
        else:
            logger.info(f"📍 未检测到GPU，使用CPU模式，PyTorch模型: {base_model_path}")
            return base_model_path
            
    except Exception as e:
        logger.warning(f"⚠️ GPU检测失败，使用默认模型: {str(e)}")
        return base_model_path

def main():
    """主函数"""
    logger.info("🔍 开始模型检查...")
    
    success = check_model()
    
    if success:
        logger.info("🎉 模型检查完成，所有测试通过！")
        sys.exit(0)
    else:
        logger.error("❌ 模型检查失败，请检查模型文件和依赖")
        sys.exit(1)

if __name__ == "__main__":
    main()