机器人视觉是使机器人具有视觉感知功能的系统。机器人视觉可以通过视觉传感器获取环境的二文图像，并通过视觉处理器进行分析和解释，进而转换为符号，让机器人能够辨识物体，并确定其位置。机器人视觉广义上称为机器视觉，其基本原理与计算机视觉类似。计算机视觉研究视觉感知的通用理论，研究视觉过程的分层信息表示和视觉处理各功能模块的计算方法。而机器视觉侧重于研究以应用为背景的专用视觉系统，只提供对执行某一特定任务相关的景物描述。机器人视觉硬件主要包括图像获取和视觉处理两部分，而图像获取由照明系统、视觉传感器、模拟－数字转换器和帧存储器等组成。根据功能不同，机器人视觉可分为视觉检验和视觉引导两种，广泛应用于电子、汽车、机械等工业部门和医学、军事领域。7968
道路车辆检测系统是指利用图像处理以及识别技术对提供的图片序列进行相关处理，找出相关的车道线。本文主要阐述了道路边缘线检测的实现过程。
在现实生活中，由于人的注意力不集中、疲劳等因素造成车祸的例子数不胜数，给社会造成了巨大的损失。车道线检测系统可以帮助驾驶员进行辅助驾驶。同时，该系统与很多学科有着密切关系，如计算机科学、通信、人工智能、信号处理、模式识别、控制理论等。因此，研究该系统具有非常实际的意义。
关键词：车道检测，图像处理，HOUGH变换，Sobel边缘梯度算子，沈俊算子，车辆检测。
Title ： Robot vision algorithms under the QNX
Abstract
The robot vision system is a system to make a robot have visual perception.
Robot vision could get the two-dimensional images of the environment by vision sensors and analysis them by visual processors to translate them to symbols in order to make the robot recognize objects and their location. Robot vision broadly known as machine vision and its basic principles are similar to the computer vision. Computer vision usually studies the general theory of the visual perception, the hierarchical information of the processes and methods to calculate the functional modules. The robot vision focuses on the use of the visual system and only provides a description of the special scene. Robot vision hardware includes image acquisition and visual processing. Image acquisition consists of lighting systems, vision sensors, analog-digital converter and frame memory, etc. According to different functions, robot vision can be pided into a visual inspection and vision to guide. They are widely used in electronics, automotive, machinery and other industrial sectors and medical and military fields.
Road vehicle detection system could analysis a serious of pictures by image processing and recognition to identify the lines and the cars.
This article focuses on the realization of the road edge line detection.
In real life, there are lots of car accidents because of people’s lack of concentration, fatigue and other factors. It caused huge losses. The system described in this article could help the driver to avoid the accidents. The system is related to many disciplines such as computer science, communications, artificial intelligence, signal processing, pattern recognition, control theory. Therefore, the system has a very real sense.
Keywords: lane detection, image processing, HOUGH transform, Sobel edge gradient operator in Anhui operator, vehicle detection
目录
1 绪论 5
1.1嵌入式系统概述 5
1.2 QNX的简介 5
1.3道路车辆检测系统概述 7
1.4道路车辆检测的研究现状 8
1.5道路车辆检测系统的应用及意义 8
1.6本文主要进行的工作 8 基于嵌入式计算平台的机器人视觉信息处理与编程:http://www.youerw.com/jisuanji/lunwen_6130.html