童鞋论文网专业提供代写本科论文、代写硕士论文、代写职称论文、代写毕业论文、论文发表等服务,上万成功案例,欢迎咨询。 网站地图  |  加入收藏  |  设为首页
童鞋论文网
热门搜索
行政 教育 营销 法律 经济 护理 心理 会计 文学 工商 计算机 旅游
站内搜索:
童鞋论文网
 当前位置:童鞋论文网 > 毕业论文 > 文章正文
 
基于视频图像处理的轨道交通客流检测方法的研究
论文作者:童鞋论文网  论文来源:www.txlunwenw.com  发布时间:2019/9/4 8:10:33  

摘要:随着城市化进程的加快,城市人口不断增长,城市的公共基础设施也在不断发展。为了解决城市人口拥堵的问题,在交通运输方面,许多城市开始发展轨道交通。轨道交通的迅速发展也带来了一些安全隐患,近几年来地铁安全事故频繁发生,大多都是因为对客流信息了解的不全面,不能及时进行客流预警,因此本文主要是通过视频图像处理的技术来对轨道交通的客流检测方法进行研究通过将视频监控技术与卷积神经网络相结合的方法来对客流检测模型进行探讨。论文的主要工作如下:

(1)对行人检测数据库和行人检测方式进行了研究,确定了用倾斜角度拍摄的视频监控录像来对行人人头进行采集,并且通过Spyder软件的脚本程序对采集到的行人人头进行截取和制作标签,最终完成实验所需要的数据库 。

(2)利用TensorFlow的深度学习框架设计实验来验证卷积神经网络在客流检测中应用的可行性。主要包括3个方面的实验:首先将不同的特征提取网络和目标检测分类网络进行组合来探究最合适的客流检测模型;然后研究了这些模型在不同平台上的检测效率,并选择最合适的实验平台;最后将选择出来的最优模型应用到实际场景中,通过不同场景下模型的准确率和实时性来验证基于卷积神经网络的客流检测模型的可操作性。

实验结果表明:(1)Faster R-CNN的特征提取网络和Inception V2的目标检测分类网络相组合的客流检测模型的检测准确率较高,检测速度较快,基本满足客流检测模型实时可操作性的要求。(2)使用服务器平台比使用计算机单机的检测速度快,建议在进行客流检测的时候使用服务器平台。(3)Faster R-CNN的特征提取网络和Inception V2的目标检测分类网络相组合的客流检测模型 可以适用于不同场景下的客流检测,因此基于卷积神经网络的客流检测方法可行。

As the urbanization process accelerates and the urban population continues to grow, the city's public infrastructure is also evolving. In order to solve the problem of urban population congestion, many cities have begun to develop rail transit in terms of transportation. The rapid development of rail transit has also brought some safety hazards. In recent years, subway safety accidents have occurred frequently. Most of them are due to the incomplete understanding of passenger flow information. And people cannot know the passenger flow warning timely. Therefore, this paper mainly uses video image processing technology to do research. In order to study the passenger flow detection method of rail transit, the passenger flow detection model is discussed by combining video surveillance technology with convolutional neural network. The main work of the thesis is as follows:

(1) Research on the pedestrian detection database and pedestrian detection method, and determine the video surveillance video taken by the oblique angle to collect the pedestrian head, and intercept and mark the collected pedestrian head through the script program of Spyder software. And finally complete the database needed for the experiment.

(2) Using TensorFlow's deep learning framework design experiments to verify the feasibility of convolutional neural networks in passenger flow detection. It mainly includes three aspects of experiment: firstly, different feature extraction networks and target detection classification networks are combined to explore the most suitable passenger flow detection model; then the detection efficiency of these models on different platforms is studied, and the most suitable experiment is selected. Finally, the selected optimal model is applied to the actual scene, and the operability of the passenger flow detection model based on convolutional neural network is verified by the accuracy and real-time of the model under different scenarios.

The experimental results show that: (1) The traffic detection network combined with the feature extraction network of Faster R-CNN and the target detection classification network of Inception V2 has higher detection accuracy and faster detection speed, which basically meets the real-time operability of passenger flow detection model. Requirements. (2) The use of the server platform is faster than the use of a computer stand-alone machine. It is recommended to use the server platform for passenger flow detection. (3) The passenger flow detection model combined with the feature extraction network of Faster R-CNN and the target detection classification network of Inception V2 can be applied to passenger flow detection in different scenarios. Therefore, the passenger flow detection method based on convolutional neural network is feasible.

关键词:视频图像处理;轨道交通;客流检测;卷积神经网络;深度学习

video image processing; rail transit; passenger flow detection;convolutional neural network;deep learning

上一篇:基于卷积神经网络的城市燃气管道故障诊断技术研究     下一篇:地铁站火灾和人员疏散仿真模拟
 
论文客服 95702844
论文热线18796337551
本站网址 www.txlunwenw.com
郑重承诺 原创,包修改,包通过!
童鞋论文网真诚欢迎新老客户的光临与惠顾!
 
 热门文章
· 住宅建设用地使用权续期研究
· 用户锁定效应视角下的网易云音
· 众筹产品创新性对融资绩效的影
· 南京市城镇居民冷鲜肉认知度问
· 摩拜共享单车可持续盈利研究
· 催生地方政府隐性债务风险的原
· 社交媒体中不实信息转发主体心
· 基于网络平台的SPA模式研究——
· 基于游客感知的云台山景区微信
· 新媒体时代背景下乡村旅游企业
 
 推荐文章
· 论商业性小屏直播的主持互动策
· 新兴偶像经济发展现状及发展建
· 辅助线在小学生空间观念培育中
· 罗浮山国家级风景名胜区探险旅
· 基于《预算法》的地方政府债务
· 商品房买卖合同“霸王条款”存
· 中泰水果贸易竞争力比较分析
· 我国企业内部转移价格应用现状
· 虚拟现实技术下的创意短片《萌
· 核定征税与查账征税实施现状及
 
 相关文章
· 环境中三氯生分布及其检测方法
· 论当代中国画应该站在传统的基
· 新木刻艺术“经纪人”鲁迅研究
· 从“人史”浅谈鲁迅与浙东学派
· 张锐锋散文集《文学王》解构
· 参考文献在毕业论文中所反映出
· 工科本科生毕业论文质量提高研
· 高职高专毕业论文质量提高探讨
· 商务英语毕业论文文献综述写作
· 体育教育专业本科毕业论文现状

论文原创,准时交稿,包修改,包通过,八年信誉,品质保证
童鞋论文网提供代写本科论文,代写专科论文,代写毕业论文,代写职称论文,代写硕士论文,论文发表及查重服务
备案号:苏ICP备11011187号-1 苏公网安备32030502000117号
版权归童鞋论文网所有 禁止一切商用盗用