当前位置:主页 > 科技论文 > 软件论文 >

机器人视觉框架及视觉原型系统的设计与实现

发布时间:2018-09-05 10:46
【摘要】:本项目应用于北京一维弦科技有限责任公司自主研究开发的机器人MoRo上,是机器人的视觉部分。视觉框架包含的模块有算法管理、消息机制、计算工厂、数据管理和进程间通讯等。视觉框架可以根据机器人的命令,正确地运行请求的视觉功能,并把最终的处理数据传递给机器人系统其它部分使用,比如肢体控制部分。算法人员利用框架提供的接口,结合算法的逻辑定义算法单元和它们之间的连接可以实现各种视觉功能,例如人形追踪、物体识别等。本人主要负责的模块是算法管理模块、消息机制模块、数据管理模块和原型系统。本人在一维弦科技有限责任公司实习期间,主要参与了视觉框架的设计、开发、单元测试、维护和版本迭代等工作。具体工作如下(1)参与讨论制定需求。根据算法组的使用需求,并分析之前架构的优点与缺点,制定了视觉框架的需求分析说明书。(2)参与并完成了视觉框架的系统设计。通过分析需求以及各种设计模式的特点完成框架的系统设计,以及参与了绘制UML类图、主要逻辑时序图并编写部分详细设计说明。(3)负责或参与的部分。算法管理模块,系统内部消息机制,数据管理模块和原型系统。(4)搭建测试架构并完成单元测试。编写测试用例,根据测试用例搭建测试架构,编写测试代码,并根据测试结果修改Bug。(5)在使用过程中的维护。在算法组同时使用过程中负责协助解决问题,完善整个框架的功能需求,使系统更加稳定。在新视觉框架加入机器人系统后,视觉功能并行数目增加,机器人系统运行效率提高。提高开发人员代码复用性,减少代码冗余。本架构设计合理,具有良好的可复用性、可扩展性,满足高内聚、低耦合的软件设计原则。
[Abstract]:This project is applied to the robot MoRo, which is developed by Beijing one dimensional string Science and Technology Co., Ltd. It is the vision part of robot. The visual framework includes modules such as algorithm management, message mechanism, computing factory, data management and inter-process communication. The visual framework can run the requested visual function correctly according to the robot's command, and transfer the final processing data to other parts of the robot system, such as the limb control part. By using the interface provided by the framework and the logical definition of the algorithm unit and the connection between them, the algorithmic personnel can realize various visual functions, such as human shape tracking, object recognition and so on. My main responsible module is algorithm management module, message mechanism module, data management module and prototype system. I took part in the design, development, unit testing, maintenance and version iteration of the visual framework during my internship in one dimensional string Science and Technology Co., Ltd. The specific work is as follows: (1) to participate in the discussion and development of needs. According to the requirements of the algorithm group and analyzing the advantages and disadvantages of the previous architecture, the requirements analysis specification of the visual framework is formulated. (2) the system design of the visual framework is completed. By analyzing the requirements and the characteristics of various design patterns, the system design of the framework is completed, as well as the drawing of UML class diagrams, the main logic sequence diagrams and the writing of some detailed design instructions. (3) the responsible part or the participating part. Algorithm management module, system internal message mechanism, data management module and prototype system. (4) build test architecture and complete unit test. The test case is written, the test structure is built according to the test case, the test code is written, and the maintenance of Bug. (5) is modified according to the test results. In the process of using the algorithm group at the same time, it is responsible to help solve the problem, improve the functional requirements of the whole framework, and make the system more stable. After the new vision frame is added to the robot system, the number of parallel visual functions increases and the efficiency of the robot system is improved. Improve developer code reuse and reduce code redundancy. The architecture is reasonable, reusable, extensible and meets the software design principle of high cohesion and low coupling.
【学位授予单位】:北京交通大学
【学位级别】:硕士
【学位授予年份】:2017
【分类号】:TP242;TP391.41

【参考文献】

相关期刊论文 前3条

1 陈天超;;单例设计模式研究[J];福建电脑;2016年08期

2 任亮;;三种设计模式在频率干扰分析软件中的应用[J];计算机与数字工程;2016年05期

3 陈北莉;;基于视觉信息的移动机器人控制系统设计[J];电子设计工程;2015年03期



本文编号:2224038

资料下载
论文发表

本文链接:https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2224038.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户94e89***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com