Real-Time Robotics
Department of Electrical and Computer Engineering (ECE) students Sihao Ding and Ying Li, under professor Yuan Zheng, teamed up with their computer science counterparts, Qiang Zhai and Dong Xuan, at Ohio State to make it happen. The results of their work are detailed in a new manuscript set for publication in the next Journal of the Pattern Recognition Society, as well as a previous paper published by IEEE called "Human Feet Tracking Guided by Locomotion Model."
Watch a short video of their robot in action HERE.
Developing machines that both track and follow subjects has required advanced research from the students. Within the robotics industry, finding ways for machines to perform both actions remains a challenge, especially in real-time. The students appear to have found the solution.
Inside Dreese Laboratories Friday, Ding flipped a switch on the back of a knee-high robot they created. As Li walked across the room, the electric wheels on the machine whirred to life, tracking her feet and shadowing along obediently.
“Right now the main function of this robot is to follow a specific person,” Li said. “The ultimate goal is for this robot to help the elderly when they need it.”
Research into human-following robots is important for future home, industrial and battlefield applications. Such robots could provide medical assistance in nursing homes, help passengers carry heavy luggage, or assist workers as they assemble large equipment. Current methods of tracking and following humans, however, remain too cumbersome.
“Vision-based techniques are widely used,” their research explains, “However, due to the close distance between human and robot, and the limitation in a camera’s field of view, only part of a human body can be observed most of the time.”
Human mimicking robots must detect a specific body part and then re-assemble the entire body form, piece-by-piece, to replicate its entire movement.
“To counter the challenges, we propose using a structured learning framework to simultaneously infer the observed body part and the motion,” they explain.
In other words, the robot is designed to track a continuous human form, which serves as the framework. As a result, the robot only needs to discern small parts of the entire form in order to identify whole body motion through a few observed sections.
In their previous IEEE paper, the students described their use of a particle-filtering framework, conducted through a human locomotion model, for increased feet tracking abilities.
When combined, their work shows a robot is able to discern a subject's feet, picking them out among the other possible "noise" around them, while using predictive models to gauge how that subject will walk and react accordingly.
To experiment with the theory, the students developed an intelligent robot, equipped with a microprocessor powered by two 24-volt batteries. A camera is mounted on the robot for video recording. The machine is programmed to follow a person at a 0.5 to 2.5-meter distance. The person could be walking, running, or turning left or right, as typically performed.
Their research work is supported by National Science Foundation funding.
About the research authors:
Sihao Ding received the B.S. degree from the Department of Electronic Engineering at University of Electronic Science and Technology of China, in 2011. He is currently working towards the Ph.D. degree in the Department of Electrical and Computer Engineering at The Ohio State University, Columbus, Ohio. His research interests include image and video processing, pattern recognition, and their multimedia applications.
Qiang Zhai received his BS degree in Information Security from Shanghai JiaoTong University in Shanghai, China in 2011. He is now pursuing Ph.D degree in Department of Computer Science and Engineering at The Ohio State University. His research interest is cyber-physic system and multisensor network.
Ying Li received the B.S. degree from the School of Hydropower and Information Engineering, Huazhong University of Science and Technology, Wuhan, China, in 2012. She is currently working toward the Ph.D. degree in the Department of Electrical and Computer Engineering, The Ohio State University, Columbus. Her research interests include video processing and machine learning.
Junda Zhu received the Ph.D. degree in Electrical and Computer Engineering from The Ohio State University, Columbus, Ohio in 2012. He received the B.S. and M.S. degrees from the Department of Information Science and Electronics Engineering, Zhejiang University, China. Currently, he is an assistant professor in the Department of Electrical and Computer Engineering, University of Macau. His research interests are in the field of image and video processing.
Yuan F. Zheng received the B.S. degree from Tsinghua University, Beijing, China, in 1970 and the M.S. and Ph.D. degrees in electrical engineering from The Ohio State University (OSU), Columbus, Ohio in 1980 and 1984, respectively. From 1984 to 1989, he was with the Department of Electrical and Computer Engineering at Clemson University, Clemson, SC. Currently, he is the Winbigler Professor and was the Chairman of the Department of Electrical and Computer Engineering from 1993 to 2004 at OSU, where he has been since 1989. His research interests include image and video processing for compression, object classification, object tracking, and robotics for which his current activities are in robotics and automation for high-throughput applications in biology studies. His research has been supported by the National Science Foundation, Air Force Research Laboratory, Office of Naval Research, Department of Energy, DAGSI, and ITEC-Ohio.
Dong Xuan received the BS and MS degrees in electronic engineering from Shanghai Jiao Tong University (SJTU), China, and the PhD degree in computer engineering from Texas A & M University. Currently, he is a professor with the Department of Computer Science and Engineering, Ohio State University (OSU), Columbus. His research interests include distributed computing, computer networks, and cyberspace security.
