基于DSP的电缆视频测速系统

VIP免费

3.0 牛悦 2024-11-19 10 4 1.51MB 59 页 15积分

侵权投诉

摘要

出线速度是电缆成缆机的重要参数，该值的精确度与电缆质量密切相关。因

此，精确测量电缆出线速度具有重要意义。传统的测量设备是计米器，计米器通

过测量单位时间内电缆的运行长度对速度进行间接测量。计米器是接触式测量设

备，具有精度低，实时性差，易受环境干扰等缺点。工业环境下操作环境复杂，

干扰大，计米器的测量精度受到严重限制，因此迫切需要开发一种非接触式的、

鲁棒性强的高精度实时测速系统。

视频跟踪是最近几年的新兴课题，也是机器视觉领域的研究热点。它是图像

处理、跟踪与控制等多学科的综合，在智能交通、安全监控、医疗图像处理等方

面有着广泛的应用前景。视频跟踪算法较为复杂，因此早期仅限于理论研究与计

算机仿真。最近20年，随着半导体技术的发展，特别是高性能DSP（Digital Signal

Processor，数字信号处理器）芯片的问世，由于其强大的运算能力适应了海量图

像数据高速处理的要求，使得制约数字图像处理技术发展的高实时性和大吞吐量

问题得到解决，推动着数字图像处理技术向着更深入的方向发展。基于DSP的视

频测速系统采用高性能DSP处理器作为硬件平台，集成图像采集、数据处理、逻

辑控制和数字显示等模块形成嵌入式系统，实现图像的采集、识别和目标测速，

具有结构紧凑，环境适应性强，不易受外部磁场的干扰，安装使用简单等特点。

本系统采用TI公司的TMS320DM6446作为主处理芯片，利用DSP芯片强大的数字

信号处理能力实现实时电缆识别与测速，并可实现电缆标识字符识别。基于DSP

的电缆实时测速系统，融合了DSP系统应用、光学字符识别、运动目标检测与跟

踪等多项技术，具有工作稳定性高，环境适应力强，可移动性好等优点，具有较

高的实用价值较和广阔的市场前景。

本论文开发的电缆视频测速系统具有如下几项功能：

（1）采用矩估计的简便方法实现复杂环境下的电缆实时测速，测量时间短，

精度高，模拟计算得到的计算精度可达到4%；

（2）采用小波分析的方法实现字母和数字识别，可以在高精度速度测量的同

时实现对电缆标识字符的精确识别；

（3）使用业界图像处理功能最强的DaVinci系列DSP芯片进行数据处理，整

个系统结构小巧，便于安装，并且稳定性高，实时性强；

本课题通过将视频跟踪和小波变换算法引入电缆测速领域，设计了一款基于

DSP平台的电缆速度测量系统。实现了复杂环境下电缆出线速度的非接触式快速、

精确测量，并可完成电缆标识字符的精确识别。

关键词：视频跟踪小波分析矩估计 DSP 光学字符识别

ABSTRACT

The velocity of the cable is an important parameter during it is producing from the

cabling machine. The accuracy of the velocity has a closely relationship with the

quality of cables. As a result, it is critical to measure the cable velocity. The traditional

measurment equipment is the Meters, which measuring the speed of the cable

indirectly by the length meterage. Meters are kinds of contact measure equipment,

which has some drawbacks, including low measurement accuracy, poor real-time

performance and vulnerable to environmental interference. However, Meters'

measuring accuracy is severely limited in the industrial environments because of the

complex background environment and the great interference. It is need urgently to

develop a non-contact, robust, high-precision and real-time velocity measurement

system.

Video tracking is an emerging issue in recent years, which is the focus in the field

of machine vision. It belongs to the image processing, target tracking, and many other

cross-disciplinary. In addition to military applications, it also has a wide range of

applications in intelligent transportating, satellite monitoring, security surveilling and

medical image processing system. With the development of semiconductor technology

and the advent of high-performance DSP (Digital Signal Processor) chip in the last 20

years, the digital image processor adapted to the massive high-speed image data

processing requirements, and the image processing technology had great changes,

which developed direction toward more in-depth.

DSP-based video velocity system uses high-speed DSP chip as the hardware

platform, integrating the image acquisition module, logic control module, and digital

display circuit module to form an embedded system to achieve the function of velocity

measurement, which is compact, environmental adaptability, less susceptible to

interference from external magnetic fields and simple installment. TI's

TMS320DM6446 chip is chosed in this work to achieve the prupose of velocity

measurement and character recognition. Integrating of the embedded DSP system,

optical character recognition (OCR), video speed measuement and many other

technicals, DSP-based real-time cable velocity measurement system works with high

stability, environmental adaptability and mobility, which has a high practical value

and broad market prospects.

In this paper, the development of the cable video velocity system has several

features as follows:

(1) A simple method of moment estimation is used to achieve the real-time cable

velocity measurement, which has shorter time and higher precision. Simulation

accuracy can be up to 4%;

(2) Wavelet analysis is used to identify the letters and numbers on the surface of

the cable. The speed   measurement as well as the character recgnization can be

taken at the same time under a high accuracy;

(3) The most powerful image processor-the DaVinci family of DSP chips is used

for data processing, which makes the system compact, convenient for installing and

stable;

In this paper, the video tracking and the wavelet analysis is introduced to the field

of cable velocity measurement and character recognition. A measurement system is

designed based on the DSP platform, which can be used to measure the speed of the

cable producing from the cabling machine, as well as recognize the characters on the

cable surface.

Key Words: video tracking, wavelet analysis, moment estimation, DSP,

optical character recognition

摘要

ABSTRACT

第一章绪论………………………………..……………………………..………...…1

§1.1 选题背景及研究意义…………………………………...….……………..…1

§1.2 国内外研究现状………………..………….………….…….……….………3

§1.2.1 运动目标检测技术…………………………………...…………..……3

§1.2.2 光学字符识别技术…………………………………...…………..……4

§1.3 本文成果…………………………………………….……...……….….........4

§1.3.1 本文主要工作…………………..…….......................…….…..……...4

§1.3.2 章节安排……….…..…….……...……...………………………..…..5

第二章电缆测速系统总体设计………….……………………………..………...…7

§2.1 总体方案…………………………..……...………………………......….......7

§2.2 硬件系统………………………..……...…………………………...…....…..8

§2.2.1 照明系统………….…...………….…..……...……………………….8

§2.2.2 摄像机选择……...…..………………..……...…………………...…..9

§2.3 软件系统………………………..……...…………………….……....…......10

第三章电缆图像的预处理技术………..…...………………..……………........…..12

§3.1 电缆图像的特点…………………..……...………………………......….....12

§3.2 图像坐标系统…………………..……...…………………………...…..…..13

§3.3 彩色模型转换…………………..……...…………………………...…..…..13

§3.4 电缆图像的中值滤波降噪..……....…………….…….…….…….….....….14

§3.5 灰度变换处理…………….……..……...…………………………....……..15

第四章基于小波分析的光学字符识别算法………………..…..……..........……...18

§4.1 光学字符识别概述…………………..……...……………………….....…..18

§4.2 常用的字符识别算法…………………..……...…………………...……....19

§4.2.1 像素匹配识别算法…………………..……...……………………….19

§4.2.2 结构匹配识别算法…………………..……...…………………...…..20

§4.2.3 人工神经网络识别算法…………………..……...……………...…..20

§4.3 基于小波分析的电缆字符识别…………………..……...……………..….21

§4.3.1 Haar 变换…………………..……...…………………….……….…...21

§4.3.2 多分辨率展开…………………..……...……………………......…...21

§4.3.3 一维小波变换…………………..……...……………………......…...22

§4.3.4 二维小波变换在电缆字符识别中的应用.…....…………….…........23

§4.4 电缆字符识别的算法实现……………..……...………………….........…..25

§4.4.1 算法流程…………………..……...…………….……………......…..25

§4.4.2 结果分析…………………..……...……………………….…......…..27

第五章运动目标检测与电缆测速…………………..……...…………….….....…..32

§5.1 运动目标检测方法概述………….…………..……...………………...…..32

§5.1.1 帧间差分法…………...……….…...……………………………......32

§5.1.2 背景差分法………..……………..……...…………………........…..33

§5.1.3 光流场法…………..…………..……...…………………………......34

§5.2 基于矩估计的电缆测速算法……….……………..…….....………….…..35

§5.2.1 字符区域分割……......……………..……...…………..…..........…..37

§5.2.2 特征点获取与像素速度计算.……..….....…….….…...…..........…..38

§5.2.3 实际速度换算……...…...…………..……...……………….…...…..39

第六章电缆测速系统的 DSP 实现…………………..……...……….…....…....…..42

§6.1 电缆视频测速系统概述…………….…..…....……...…………...…...…...42

§6.1.1 系统简介………………..….…..…….…….……….…………...….42

§6.1.2 DSP 用于电缆视频测速系统的优势………..………..……..….......42

§6.2 电缆视频测速系统的 DaVinci 平台实现……….……..………………….43

§6.2.1 TMS320DM6446 处理器简述…………….…....…...…………..…..43

§6.2.2 系统架构与工作流程…...…………….….………...………......…...45

第七章结束语………………………………………………..……...……….....…...48

§7.1 工作总结………………..……...……………...………………...….....…...48

§7.2 前景展望………………..……..……………………………………...…....48

参考文献………………..……...…………………………………………..…..….......50

在读期间公开发表的论文和承担科研项目及取得成果………….……..…...……..54

致谢……..………..……...……………………………………………………...........55

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

15 积分 4人已下载

立即下载 VIP免费下载

摘要：

摘要出线速度是电缆成缆机的重要参数，该值的精确度与电缆质量密切相关。因此，精确测量电缆出线速度具有重要意义。传统的测量设备是计米器，计米器通过测量单位时间内电缆的运行长度对速度进行间接测量。计米器是接触式测量设备，具有精度低，实时性差，易受环境干扰等缺点。工业环境下操作环境复杂，干扰大，计米器的测量精度受到严重限制，因此迫切需要开发一种非接触式的、鲁棒性强的高精度实时测速系统。视频跟踪是最近几年的新兴课题，也是机器视觉领域的研究热点。它是图像处理、跟踪与控制等多学科的综合，在智能交通、安全监控、医疗图像处理等方面有着广泛的应用前景。视频跟踪算法较为复杂，因此早期仅限于理论研究与计算机仿真。最近...

展开>> 收起<<

基于DSP的电缆视频测速系统.pdf

共59页,预览6页

还剩页未读，继续阅读

基于DSP的电缆视频测速系统

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

推荐作者

热门标签

举报选择: