2008年英特尔杯大学生电子设计竞赛—嵌入式系统专题邀请赛，一等奖作品：
参赛学校 杭州电子科技大学
指导教师 申兴发
参赛队员 张斌吴积冰陈宁
作品题目
(中英文对照)嵌入式自主跟踪系统-球童机器人
Embedded Autonomous Tracking System -Caddie Robot
作品
简介
随着人工智能和计算机视觉技术的发展与成熟，服务型机器人已出现在各个领域。由于高尔夫独特的运动方式和复杂的场地条件，因而对高尔夫球场的球童机器人研究具有一定的价值。目前，国内外已有一些厂商推出了相应的产品，这些产品大多应用GPS导航遥控或是无线射频控制技术，取得了一定的成功。
本作品基于Intel双核处理器的开发平台，综合运用RF无线通信，超声测距，计算机视觉，模式识别等技术，实现了机器人对运动目标的识别与跟踪。当前，国内外对静止背景下的目标跟踪，已经有大量的成熟的研究成果，但是对于前景和背景都是运动情况下的目标跟踪的研究还处于起步阶段，尚没有成熟的技术来解决此问题。本作品是对计算机视觉跟踪技术的一次大胆应用，采用多传感器协同工作的方式，尝试解决摄像头运动情况下的目标跟踪，取得了不错的效果。
为了充分发挥开发平台的双核优势，本作品中还加入了基于OpenMP标准的双核多线程优化技术，使作品的鲁棒性和实时性得到了提高。另外，本作品的研究成果可以很容易地引用到其他领域，如智能玩具，智能监控护理，军事上的自动巡逻等等。所以有理由相信，本作品具有很好的应用前景和扩展推广价值。
With the development of the artificial intelligence and computer vision technology, service-oriented robots have appeared in various fields. Considering the unique sports style and complex venues condition of golf, there is great value in the research of caddie robot of golf. At present, a number of corresponding products are introduced to the world as a certain kind of success from domestic and foreign manufacturers, most of which are based on GPS navigation system or radio frequency remote control.
This robot is designed on the basis of Intel dual-core processor, integrating the technologies of RF wireless communication, ultrasonic ranging, computer vision, pattern recognition and so on, so that it can identify and track the moving objects. At present, there are a great many mature research results, in terms of object tracking which is in the static background. but the study of the moving object tracking when both foreground and background are moving is still in its initial stage, there is no mature technique to solve this problem . This work is an innovative application of the computer vision technology in object tracking area. Integrating information from multi-sensors, the robot solves the problem of tracking moving objects with moving cameras, which has achieved good results.
To take full advantage of the dual-core platform, the work adopts the technology of dual-core multi-threaded optimization based on OpenMP standards. So that the robustness and real-time of the work has been enhanced. In addition, our research results can be easily applied to other areas, such as smart toys, intelligent monitoring care, the military patrol, and so on. we are confident that the work will have a good prospect and great value for promotion.
本帖最后由 paulhyde 于 2014-9-15 09:01 编辑
京公网安备 11010802033920号
