Mahbubur Rahman

Mahbubur Rahman

Ph.D. Candidate

Research Assistant, at ECLIPSE Cluster
Department of Computer Science and Electrical Engineering
University of Maryland, Baltimore County
1000 Hilltop Circle, Baltimore, Maryland, 21250
Office: ITE 351


M.Sc., LIT, Pusan National University, Busan, S. Korea B.Sc., CSE, University of Dhaka, Dhaka, Bangladesh

I am a Ph.D. candidate and a research assistant in computer science at University of aryland Baltimore County, Baltimore-MD. Currently, I am working in the Eclipse cluster lab. Prior to joining Eclipse cluster lab, I was a graduate student and research assistant in Logistics Information and Technology at Pusan National University. I worked there with Dr. Bonghee Hong in the Ubiquitous Computing Database Laboratory. I completed my undergraduate in computer science, the University of Dhaka.

Research Interests

I am interested in data driven system development in computer vision. In specific, my works span on applications about computer vision-based assistive technology, sensor data to monitor human health status and apply machine learning in designing automated speech recognition applications.


Human Assisted Computer Vision

Vision-based Realtime Accessibility Issues Detection

The project aims at using machine vision to detect accessibility issues on sidewalks for helping people with visually-impairment. A prior-based framework is designed to accelerate the computer vision detection process in embedded devices, which uses a cooperation of human and machine.



Vision-based Accurate Bus Stop Localization

We target to detect and localize bus stop accurately to improve the accessibility of bus station for blind riders. A bus stop detection system is developed and implemented completely in a smartphone platform.

Poster Demo


Weed Infestation Identification

We have designed to identify weed in the crop fields in smart-home platform. We developed both iOS and android application and also have backend intelligence to reduce the cost and time of identifying invasive weeds

Poster Demo


PreSight: Enabling Real-Time Detection of Accessibility Problems on Sidewalks

Zheng Li, Mahbubur Rahman, Ryan Robucci, Nilanjan Banerjee
2017 14th Annual IEEE International Conference on Sensing, Communication, and Networking (SECON)
Conference AR: 26% BibTex Download

Smartphone-based Hierarchical Crowdsourcing for Weed Identification

Mahbubur Rahman, Brenna Blackwell, Nilanjan Banerjee, Dharmendra Saraswat
Computers and Electronics in Agriculture, Elsevier, Vol. 113, pp: 14-23, 2015
Journal Paper IF: 2.1 BibTex Download