数据集来源

开始使用目标检测检测人体，再通过宽高比值变化来判断是否摔倒，该方法在密集场景不管用，所以想着找一个摔倒数据集来进行检测是否摔倒，附上数据集来源：
中国华录杯——定向算法赛（人体摔倒姿态识别）这个训练集数据集我没有找到，官网下载的也只是未标注的测试集
Le2i Fall detection Dataset
• Annotations

标签格式为txt格式
标签前两行为跌倒帧数的起始与终止点，也存在没有这两行的文件
之后一行就是六个字段，帧编号， XX，人物bbx

Le2i Fall detection Dataset 原文下载速度不敢恭维，附上百度云下载链接, 提取码：qghr
中国华录杯——定向算法赛（人体摔倒姿态识别）测试集，无标注，自己找的一些算法打的标注，还算过得去，需要自取百度云百度云下载链接 , 提取码：asdh
以上两个数据集要是侵权了会删除

数据的处理

对Le2i Fall detection Dataset数据转换成voc格式的xml文件：

数据集解压后如下图，不知道有两个场景是没有标注文件的，号称假摔，所以我把他删了
解压后运行如下代码

# coding=utf-8
"""
@Time: 2021/3/9 10:23
@IDE: PyCharm
@Author: chengzhang
@File: processVideo.py
"""
import os
from tqdm import tqdm
import cv2
from xml.dom.minidom import Documentcount = 0def process(fileDir):dataset = [os.path.join(fileDir, i) for i in os.listdir(fileDir)]for dir in dataset:Annotation = os.path.join(dir, "Annotation_files")Annotations = [os.path.join(Annotation, i) for i in os.listdir(Annotation)]for ann in tqdm(Annotations):txt = open(ann).readlines()if len(txt[0].split(',')) < 2:videoPath = ann.replace('Annotation_files', 'Videos').replace('.txt', '.avi')start, end = int(txt[0].replace('\n', '')), int(txt[1].replace('\n', ''))video(videoPath, txt, start, end)def video(videoPath, txt, start, end):global countvs = cv2.VideoCapture(videoPath)ind = 1while True:success, frame = vs.read()if frame is None:breakif (ind + 8) >= start and ind <= end:Jpg = '{:06d}.jpg'.format(count)outPutJpg = os.path.join(save, Jpg)outPutXML = outPutJpg.replace('jpg', 'xml')cv2.imwrite(outPutJpg, frame)count += 1imgShape = frame.shapeinfo = txt[ind].replace('\n', '').split(',')_, _, xmin, ymin, xmax, ymax = infobboxsInfo = [[int(xmin), int(ymin), int(xmax), int(ymax)]]bbox2xml(bboxsInfo, imgShape, Jpg, outPutXML)ind += 1def bbox2xml(bboxsInfo, imgShape, outPutJpg, outPutXML):# for boxInfo in bboxsInfo:xmlBuilder = Document()annotation = xmlBuilder.createElement("annotation")  # 创建annotation标签xmlBuilder.appendChild(annotation)Pheight, Pwidth, Pdepth = imgShapefolder = xmlBuilder.createElement("folder")  # folder标签folderContent = xmlBuilder.createTextNode("VOC2007")folder.appendChild(folderContent)annotation.appendChild(folder)filename = xmlBuilder.createElement("filename")  # filename标签filenameContent = xmlBuilder.createTextNode(outPutJpg)filename.appendChild(filenameContent)annotation.appendChild(filename)size = xmlBuilder.createElement("size")  # size标签width = xmlBuilder.createElement("width")  # size子标签widthwidthContent = xmlBuilder.createTextNode(str(Pwidth))width.appendChild(widthContent)size.appendChild(width)height = xmlBuilder.createElement("height")  # size子标签heightheightContent = xmlBuilder.createTextNode(str(Pheight))height.appendChild(heightContent)size.appendChild(height)depth = xmlBuilder.createElement("depth")  # size子标签depthdepthContent = xmlBuilder.createTextNode(str(Pdepth))depth.appendChild(depthContent)size.appendChild(depth)annotation.appendChild(size)for bbox in bboxsInfo:label = "fall"box_xmin, box_ymin, box_xmax, box_ymax = bboxobject = xmlBuilder.createElement("object")picname = xmlBuilder.createElement("name")nameContent = xmlBuilder.createTextNode(label)picname.appendChild(nameContent)object.appendChild(picname)pose = xmlBuilder.createElement("pose")poseContent = xmlBuilder.createTextNode("Unspecified")pose.appendChild(poseContent)object.appendChild(pose)truncated = xmlBuilder.createElement("truncated")truncatedContent = xmlBuilder.createTextNode("0")truncated.appendChild(truncatedContent)object.appendChild(truncated)difficult = xmlBuilder.createElement("difficult")difficultContent = xmlBuilder.createTextNode("0")difficult.appendChild(difficultContent)object.appendChild(difficult)bndbox = xmlBuilder.createElement("bndbox")xmin = xmlBuilder.createElement("xmin")xminContent = xmlBuilder.createTextNode(str(box_xmin))xmin.appendChild(xminContent)bndbox.appendChild(xmin)ymin = xmlBuilder.createElement("ymin")yminContent = xmlBuilder.createTextNode(str(box_ymin))ymin.appendChild(yminContent)bndbox.appendChild(ymin)xmax = xmlBuilder.createElement("xmax")xmaxContent = xmlBuilder.createTextNode(str(box_xmax))xmax.appendChild(xmaxContent)bndbox.appendChild(xmax)ymax = xmlBuilder.createElement("ymax")ymaxContent = xmlBuilder.createTextNode(str(box_ymax))ymax.appendChild(ymaxContent)bndbox.appendChild(ymax)object.appendChild(bndbox)annotation.appendChild(object)f = open(outPutXML, 'w')xmlBuilder.writexml(f, indent='\t', newl='\n', addindent='\t', encoding='utf-8')f.close()if __name__ == '__main__':fileDir = r"E:\数据集\FallDataset\dataset"  # 源数据集路径save = r"E:\数据集\FallDataset\FallDataset"  # 保存voc格式路径process(fileDir)

模型训练

使用U神的yolov5，链接
将voc格式的数据集转换成yolov5所需要的格式，数据格式如下：其中txt标注为下面代码生成的归一化数据

# coding=utf-8
"""
@Time: 2020/9/17 9:07
@IDE: PyCharm
@Author: chengzhang
@File: voc2v5.py
"""
import os
import glob
from tqdm import tqdm
import xml.etree.ElementTree as ET
import shutil
import argparse
from random import shuffle
import cv2def check_x_y_w_h(norm_x, norm_y, norm_w, norm_h):norm_x = 0.00001 if norm_x <= 0 else norm_xnorm_x = 0.99999 if norm_x >= 1 else norm_xnorm_y = 0.00001 if norm_y <= 0 else norm_ynorm_y = 0.99999 if norm_y >= 1 else norm_ynorm_w = 0.00001 if norm_w <= 0 else norm_wnorm_w = 0.99999 if norm_w >= 1 else norm_wnorm_h = 0.00001 if norm_h <= 0 else norm_hnorm_h = 0.99999 if norm_h >= 1 else norm_hreturn norm_x, norm_y, norm_w, norm_hdef convert(size, box):dw = 1. / (size[0])dh = 1. / (size[1])x = (box[0] + box[1]) / 2.0 - 1y = (box[2] + box[3]) / 2.0 - 1w = box[1] - box[0]h = box[3] - box[2]x = x * dww = w * dwy = y * dhh = h * dhx, y, w, h = check_x_y_w_h(x, y, w, h)return (x, y, w, h)def write_txt(train_xml, dataset):with open(dataset, "w", encoding="utf-8") as f:for xml in tqdm(train_xml):f.write(xml.replace('.xml', '.jpg') + '\n')target_label_txt = xml.replace('.xml', '.txt')img = cv2.imread(xml.replace('.xml', '.jpg'))h, w, c = img.shape# 输入文件xmlin_file = open(xml)# 输出label txtout_file = open(target_label_txt, 'w')tree = ET.parse(in_file)root = tree.getroot()for obj in root.iter('object'):try:cls = obj.find('name').textexcept:continueif cls not in classes:continuexmlbox = obj.find('bndbox')b = (float(xmlbox.find('xmin').text), float(xmlbox.find('xmax').text), float(xmlbox.find('ymin').text),float(xmlbox.find('ymax').text))bb = convert((w, h), b)out_file.write(str(0) + " " + " ".join([str(a) for a in bb]) + '\n')def check_dir(dir):try:shutil.rmtree(dir)except:os.makedirs(dir)if not os.path.exists(dir):os.makedirs(dir)def parse_annotation(img_dir):all_imgs = []all_xmls = []for tmpdir in img_dir:for home, dirs, files in os.walk(tmpdir):for filename in files:# 文件名列表，包含完整路径if filename.endswith((".jpg", ".jpeg", ".png")):filePath = os.path.join(home, filename)all_imgs.append(filePath)elif filename.endswith(".xml"):filePath = os.path.join(home, filename)all_xmls.append(filePath)return all_xmlsif __name__ == '__main__':classes = ['fall']parser = argparse.ArgumentParser()parser.add_argument("--train_txt", default="/data/share/imageAlgorithm/zhangcheng/code/yolov5/data/train_fall.txt")parser.add_argument("--test_txt", default="/data/share/imageAlgorithm/zhangcheng/code/yolov5/data/val_fall.txt")flags = parser.parse_args()imgPathList = ["/data/share/imageAlgorithm/zhangcheng/dataset/falldataset/FallDataset/","/data/share/imageAlgorithm/zhangcheng/dataset/falldataset/FallDataset_A/"]dataset = parse_annotation(imgPathList)shuffle(dataset)num = len(dataset)trainNum = int(0.9 * num)trainList = dataset[: trainNum]testList = dataset[trainNum:]write_txt(trainList, flags.train_txt)write_txt(testList, flags.test_txt)

根据上面处理的数据进行训练

CUDA_VISIBLE_DEVICES=4,5,6,7 nohup python train.py --img 640 --batch 16 --epochs 600 --data ./data/fallDown.yaml --weights ./weights/yolov5l.pt > train_fall.log 2>&1 &

训练曲线如下：
30个epoch结果

摔倒检测+yolov5相关推荐

Python+Yolov5跌倒检测摔倒检测人物目标行为人体特征识别
Python+Yolov5跌倒检测摔倒检测人物目标行为人体特征识别如需安装运行环境或远程调试,见文章底部个人QQ名片,由专业技术人员远程协助! 前言这篇博客针对<<Python+ ...
用AidLux跑老人摔倒检测项目，看AI助老轻松落地
据国家卫健委预计,到2035年,我国60岁及以上老年人口数量将突破4亿,在总人口中的占比将超过30%. 届时我国将进入重度老龄化阶段.面对日益严峻的老龄化趋势,AI助老.智慧养老被加速提上日程. 对于 ...
摔跤视频软件测试,Apple Watch 4摔倒检测立功：成功救人一命，网友：马上入手！...
原标题:Apple Watch 4摔倒检测立功:成功救人一命,网友:马上入手! 在2018年的苹果秋季发布会上,iPhone XS系列无疑是整场发布会的主角,不过同时推出的Apple Watch Se ...
摔倒、摔倒检测数据集
近期学习摔倒检测,接触摔倒数据集,自学笔记,仅用作个人复习. the UR fall detection dataset (URFD) the fall detection dataset (FDD) ...
目标检测 YOLOv5 - 模型的样子
目标检测 YOLOv5 - 模型的样子 flyfish 文章目录目标检测 YOLOv5 - 模型的样子开始加载模型文件模型的层模型的属性模块的名称以及模块本身模型的权重模型权重的名字和权 ...
目标检测 YOLOv5 - ncnn模型的加密 C++实现封装库和Android调用库示例
目标检测 YOLOv5 - ncnn模型的加密 C++实现封装库和Android调用库示例 flyfish 文章目录目标检测 YOLOv5 - ncnn模型的加密 C++实现封装库和Android调 ...
《深度学习与目标检测 YOLOv5》
<深度学习与目标检测 YOLOv5> flyfish 基础深度学习基础 - 向量深度学习基础 - 累加符号和连乘符号深度学习基础 - 最大似然估计深度学习基础 - 朴素贝叶斯深度 ...
计算机视觉摔倒检测,有关摔倒检测数据集（fall detection databases）
近期学习摔倒检测,接触摔倒数据集,自学笔记,仅用作个人复习. the UR fall detection dataset (URFD) the fall detection dataset (FDD) ...
姿态检测树莓派_基于深度学习的树莓派老人摔倒检测系统的制作方法
本发明属于电子器件技术领域,涉及基于深度学习的树莓派老人摔倒检测系统. 背景技术: 日益庞大的老年群体已成为人们关注的焦点.由于老年人身体活动不便等特点,摔倒已成为我国人员伤亡的第四位原因,而意外摔倒 ...

摔倒检测+yolov5

摔倒检测

数据集来源

数据的处理

模型训练

摔倒检测+yolov5相关推荐

最新文章

热门文章