Faiz M. Hasanuzzaman's REU [2011] Web Page

About Me

Email: faiz.zaman35@gmail.com
Office: The Center for Computational Biomedicine Imaging and Modeling (CBIM)
Home Institution: City College of New York The Grove School of Engineering
Project: Identifying Objects by Incorporating RFID

About My Project

Combining computer vision system and Radio Frequency to increase the accuracy rate for detecting medicine bottles.

My project is about helping elderly people to locate there medicine bottles. The goal of the summer research project is to use computer vision algorithms and radio frequency tags to localize the objects. To accomplish this, I will use RFID tags on each bottle which will act like a sensor. Then the camera can detect weather or not if any of the bottles are within the field.

Research Log

Week 1

I spoke to my mentor Dr. Liu who is researcher at the computer science dept. I received information about my project including the hardware and software to use. This project is mostly designing a system to assist elderly people. My other mentor Dr. Tian who is not here in Rutgers but communicates with us through email and Skype has given the hardware which is the RFID's

Week 2

This week, I tested the RFID's system and the tags on simple cases. Such cases were embedding the tags within the bottle so it can detect how bottles or tags are there. In this case the tags can be also referred to bottles. Then I designed a simulation to link the hardware to the software using C in Visual Studio 2010. I wrote a program that can easily tell you what objects has been detected and what objects are out of the range.

My mentor has also introduced me to different types of algorthim that I can use for the vision system.

Week 3

This week I created an algorithm that can locate objects that are moving. This algorithm uses background subtraction method to see what objects are there in the foreground. More algorithms will be designed as the days goes on. The background subtraction method can be used to see what foreground objects are there when there is a static background. To simplify, taking an Image A which is an apple would be the background image. Then Image B is someone taking the apple. The background subtraction method can tell you only that the apple is taking and nothing else. So the Foreground Image = Image A - Image B.

Week 4

This week, I have finished my algorthim with the blog detection, and then applied color detection just to detect tags. I have not put hardware codes and the vidoe codes together.

Week 5

This week, my mentor suggested me if instead using color to detect the colors, we can use texture recogation. This will enhance and will better to detect the right tags rather then just detecting only the ones that are blue. Also I have to write a conference paper at the IEEE Bio Medince paper.

Week 6

This week, I got results for my project. It detected 100% for getting the RFID tags. Also got results for background subtraction.

Week 7

This week, I am writing the conference paper.

Week 8

Almost finishing the paper and got new results. Also writing a conference paper to IEEE Bio Medicine.

References & Links

Here are the papers I have read for my project:
  1. An Improved Adaptive Background Mixture Model for Real time Tracking with Shadow Detection., P. KaewTraKulPong and R.Bowden - Vision and Virtual Reality Group.
  2. GPU implementation of Extended Gaussian mixture model for Background subtraction, Vu Pham, Phong Vo, Vu Thanh Hung, Le Hoai Bac - Link.
  3. An efficient Video Segmentation Algorithim with Real time Adaptive Threshold Technique, Yasira Beevi C P, Dr. S. Natarajan - Link.
  4. Object Detection Combining Recognition and Segmentation, Liming Wang, Jianbo Shi, Gang Song, I-fan Shen - Link.
  5. Algorithms for Video Detection and Segmentation with Application to Content-Based Multimedia Systems, Huitao Luo - Link.
Here are my mentor's two websites, and the REU website: