旋翼式水表图像识别自动抄收系统和移动端应用设计

发布时间：2018-03-11 12:22

本文选题：图像识别　切入点：机器视觉　出处：《湖北工业大学》2017年硕士论文　论文类型：学位论文

【摘要】：供水计量是水司运营中最重要的环节,在二次供水和一户一表的政策推行之后,机械水表抄表任务剧增,人工上门抄读工作量大,抄表准确度低,许多室内水表难于抄见,抄表人员也无法监管。而且上门抄表,也有安全隐患。智能远传水表发展多年,因为价格高,维护难,还存在电源问题、机电同步问题,水司使用起来困难比较多。因此目前市面上使用中的80%以上还是机械水表,而机械水表的抄表效率低下,极为不适应信息化的需求。在智慧城市建设中,对机械水表进行快速便捷的智能化改造,尤为重要。因此,本文采用了一种快捷简便的方式,在不更改任何机械水表部件的前提下,完成机械水表的智能化,实现远程抄表。该系统只需要在原有的水表上安装一个摄像模块,成本低廉,实现简便,准确安全。该方案中,水表计量与采集与通讯相互分离,各司其职,各自保证高效能,使整个系统快速、安全,节能,高效。实践证明,通过此设计,机械水表很方便就可以转换为物联网水表,成为智慧计量的传感器。本文基于机器视觉和图像识别,采取了两种方式完成机械水表抄表任务。一种方式是以手机拍照来取得水表读数,通过云端识别水表读数。另一种方式是在机械水表上直接安装摄像模块,完成图像采集,通过云端识别来读取水表读数,并存储现场图像,便于抄表和管理,也便于核查比对。旋翼式多流束水表显示部分一般是字轮式,字轮的数字读取系统的设计主要有:光学成像模块设计,图像采集、图像分析、模数转换模块设计、人机接口模块设计、集中器设计和移动端APP设计。该系统已经过硬件设计和软件程序嵌入,实现了固定式和移动式两种抄表功能。本文首先介绍系统的整体设计思想和设计相关要点,然后分章节具体介绍各功能模块的实现方法。
[Abstract]:The metering of water supply is the most important link in the operation of the Water Department. After the implementation of the policy of secondary water supply and one meter per household, the task of mechanical water meter reading has been greatly increased, the workload of manual door-to-door reading is large, the accuracy of meter reading is low, and many indoor water meters are difficult to copy. Meter reading personnel can not be supervised. Moreover, there are hidden dangers to safety when reading meters at home. Intelligent remote water meters have developed for many years because of high prices, difficult maintenance, problems with power supply and electromechanical synchronization. It is more difficult for the Water Division to use. Therefore, more than 80% meters in use on the market at present are still mechanical water meters, and the reading efficiency of mechanical water meters is so low that they are extremely unsuitable for the needs of information. In the construction of intelligent cities, It is very important to carry on the quick and convenient intelligent transformation to the mechanical water meter. Therefore, this paper adopts a quick and simple way to accomplish the intellectualization of the mechanical water meter without changing any parts of the mechanical water meter. The system only needs to install a camera module on the original water meter, which is cheap, simple, accurate and safe. In this scheme, the metering and collecting of water meter are separated from each other, their respective duties are performed, and their respective high efficiency is guaranteed. It is proved by practice that the mechanical water meter can be easily converted into an Internet of things water meter and become a smart metering sensor. This paper is based on machine vision and image recognition. Two methods are adopted to accomplish the task of reading the water meter by mechanical meter. One is to take pictures of the water meter by mobile phone to get the water meter reading, and the other is to directly install the camera module on the mechanical water meter to complete the image collection. Through cloud recognition to read water meter readings, and store field images, easy to meter reading and management, but also easy to check and compare. Rotary-wing multi-flow beam water meter display part is generally word wheel, The design of digital reading system of word wheel mainly includes: optical imaging module design, image acquisition module, image analysis module design, analog to digital conversion module design, man-machine interface module design. Design of concentrator and APP of mobile terminal. The system has been embedded in hardware design and software program, realizing two functions of fixed and mobile meter reading. This paper first introduces the whole design idea of the system and the key points of the design. Then it introduces the realization method of each function module in detail.
【学位授予单位】：湖北工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TU991.62

【参考文献】