RESEARCH GROUPS DETAIL

Detection and Recognition for evolutionary robotics (DRER) Research Group

Introduction

According to Stefano Nolfi, “the Evolutionary robotics is inspired by the Darwinian principle of selective reproduction of the fittest, it views robots as autonomous artificial organisms that develop their own skills in close interaction with the environment and without human intervention. Drawing heavily on biology and ethology, it uses the tools of neural networks, genetic algorithms, dynamic systems, and biomorphic engineering.” The main aim of DRER is to concentrate on the concept of computer vision employed by evolutionary robotics for areas such as surveillance, crowd security and target tracking.

The main constraint in any robotics vision algorithm is the development of decision criteria and making it illumination invariant. As the group leader of DRER my research interests are mainly focused on employing pattern recognition algorithms which are invariant to illumination changes for defense and security applications. During my PhD at University of Sussex I proposed a correlation filter named SPOT-MACH which is an abbreviation for Spatial Domain Optimal Trade-Off Maximum Average Correlation Height Filter. The proposed SPOT-MACH can be used to maximize the height of the correlation peak in the presence of distortion and provides resistance to background clutter. A local normalization of the image using a fixed size kernel ensures in the case of a large illumination gradient that there are no false detections. Recently I was invited to present a paper in a SPIE conference in Baltimore, USA which is the largest defense symposium in the world, comparing the proposed technique to a popular feature transformation technique considered to be extremely robust and reliable. Currently also in collaboration with the evolutionary robotics group at University of Sussex to develop genetic algorithms for controlling a robot using vision markers which is an industrially funded project. Also have good links with Kuwait Ministry of Defense and have collaborated on their border security systems in the past.

Group Name: Detec tion and Recognition for evolutionary robotics (DRER)
Team Lead: Dr. Akber Abid Gardezi
Research Area:

Pattern Recognition, Embedded Systems, Correlation Filters, Evolutionary Robotics

Group Members

Dr. Akber Abid Gardezi
Mr. Hafiz Hassaan Saeed

List of Projects:

• USB 2.0 Protocol Engine, Controller, External Interface design and implementation on FPGA
• Design and fabrication of application development system across MC68VZ328
• Design and fabrication of a PCB board with a Digital Camera Interface
• Design and fabrication of a PCB board with Voice over IP interface
• Mobile robot employing artificial vision for route navigation using Lego robot
• GPS chip interfacing using ARDUINO MEGA 2560 with SD-CARD Interface
• Automated Bartender, A fully functional mechatronic assembly (Project Supervisor)
• Intelligent Maze Solver using Arduino Robot + ELISA 3 Swarm robots (Project Supervisor)
• Intelligent temperature control using FPGA (Project Supervisor)