将yolov5中val.py中save_txt函数改为根据置信度大小不同存入不同的excel

好的，你可以将以下代码替换val.py中的save_txt函数，实现根据置信度大小将结果存入不同的excel文件中：

def save_txt(save_dir, img, detections, conf_thres=0.5):
    # Create directory
    if not os.path.exists(save_dir):
        os.makedirs(save_dir)

    # Get image size
    img_size = img.shape[2]

    # Create excel files for different confidence levels
    high_conf_f = openpyxl.Workbook()
    high_conf_sheet = high_conf_f.active
    high_conf_sheet.append(['class', 'confidence', 'xmin', 'ymin', 'xmax', 'ymax'])

    med_conf_f = openpyxl.Workbook()
    med_conf_sheet = med_conf_f.active
    med_conf_sheet.append(['class', 'confidence', 'xmin', 'ymin', 'xmax', 'ymax'])

    low_conf_f = openpyxl.Workbook()
    low_conf_sheet = low_conf_f.active
    low_conf_sheet.append(['class', 'confidence', 'xmin', 'ymin', 'xmax', 'ymax'])

    # Iterate through detections and save to the appropriate excel file based on confidence level
    for x1, y1, x2, y2, conf, cls_conf, cls_pred in detections:
        if cls_conf > conf_thres:
            if cls_conf >= 0.9:
                high_conf_sheet.append([classes[int(cls_pred)], cls_conf.item(), x1.item(), y1.item(), x2.item(), y2.item()])
            elif cls_conf >= 0.5:
                med_conf_sheet.append([classes[int(cls_pred)], cls_conf.item(), x1.item(), y1.item(), x2.item(), y2.item()])
            else:
                low_conf_sheet.append([classes[int(cls_pred)], cls_conf.item(), x1.item(), y1.item(), x2.item(), y2.item()])

    # Save excel files
    high_conf_f.save(os.path.join(save_dir, 'high_conf.xlsx'))
    med_conf_f.save(os.path.join(save_dir, 'med_conf.xlsx'))
    low_conf_f.save(os.path.join(save_dir, 'low_conf.xlsx'))
}

这个函数会在指定的目录中创建三个不同的excel文件，分别存储置信度高、中、低的检测结果。你可以调整`conf_thres`参数来控制置信度的阈值。记得要安装`openpyxl`库，否则无法运行。