Contents
Embedded Vision Applications
- Abstract: Embedded vision refers to machines that understand their environment through visual means. By “Embedded” we’re referring to any image-sensor-inclusive system that isn’t a general-purpose computer. A diversity of robust embedded vision processing product options exist: microprocessors and embedded controllers, application-tailored SoC’s, DSP’s, Graphic Processors, ASICs and FPGAs. An FPGA is an intriguing silicon platform for realizing embedded vision, because it approximates the combination of the hardware attributes of an ASIC — high performance and low power consumption — with the flexibility and time-to-marker advantages of the software algorithm alternative running on a CPU, GPU or DSP.
- Typical Applications:
- Road Sign Detection, Pedestrian Detection and Tracking.
- Medical Applications.
- Surveillance.
- Robotics, Military drones.
- Assembly Line Inspection.
- Resources:
- Faculty advisor(s): Shawki Areibi
- Students:
Data Mining Techniques for Classification/Prediction
- Generally, data mining (sometimes called data or knowledge discovery) is the process of analyzing data from different perspectives and summarizing it into useful information – information that can be used to increase revenue, cuts costs, or both. Data mining software is one of a number of analytical tools for analyzing data. It allows users to analyze data from many different dimensions or angles, categorize it, and summarize the relationships identified. Technically, data mining is the process of finding correlations or patterns among dozens of fields in large relational databases. In this project students will attempt to design Data Mining Algorithms that can be used to extract important information from Raw Data that is Unbalanced, with high dimensionality for multi-class systems.
- Typical Applications:
- Bussiness.
- Medical Applications.
- Science and Engineering.
- Resources:
- Faculty advisor(s): Shawki Areibi
- Students:
Fingerprint Based User Authentication For Embedded Systems
- User authentication refers to the process of verifying the identity of a user. This is becoming an important security requirement in various embedded systems. This project investigates the problem of supporting efficient fingerprint-based user authentication in embedded systems. There are two types of finger-based authentication techniques: graph-based and minutaie-based. In this project students will concentrate on the latter because minutiae are widely believed to be the most discriminating and reliable features of a fingerprint.
- Typical Applications:
- Security Communication.
- Secure information storage.
- Resources:
- Faculty advisor(s): Shawki Areibi
- Students:
Face Detection/Recognition
- Given still or video images of a scene, identify or verify one or more persons in the scene using a stored database of faces. Using either (a) Statistical Methods such as template matching or projection based methods (b) Artificial Neural Networks based on Geometrical local feature based techniques or Holistic-based ANN.
- Typical Applications:
- Entertainment, Video games, Virtual reality, Human robot interaction
- Smart Cards, Driver’s licences, immigration, voter registration
- Law Enforcement, advanced video surveillance, shoplifting.
- Resources:
- Faculty advisor(s): Shawki Areibi
- Students:
Blind Sensory Device
- Blind People are unable to attain some information about their external environment. They can’t tell the colours of objects, how hot an object at a distance is or how far objects are from them. This project should explore the potential of a device that can use an embedded system (CPU/FPGA) and multiple sensors to provide information to the blind. The output to the blind should be an audio signal (text to speech) that can relay information in a private manner.
- Typical Applications:
- Aid Blind attain useful information about environment.
- Resources:
- Faculty advisor(s): Shawki Areibi
- Students:
Wireless Sensors
- The commercialization of wireless sensor networks has begun thanks to advances in hardware device miniaturization, lower power consumption levels, and small software operating systems, and wireless sensor-net technology is expected to become a pervasive element of our daily lives once certain technical kinks are ironed out. Applications once thought to be unrealistic are now achievable because of wireless sensing technology, which facilitates signal monitoring in hard-to-access locations and makes factory-floor cabling redundant. The fundamental components of wireless sensor nets are minuscule “mote” computers that run on batteries and use radio to communicate with each other as well as configure themselves into ad hoc networks. “Wireless sensor nets will become most ubiquitous in commercial markets for the near future, with applications ranging from security and bio-detection to building and home automation, industrial control, pollution monitoring, and agriculture,” says Avaak CTO Bar-Giora Goldberg. Sensors’ presence in the automotive industry is particularly strong in tire-pressure monitoring systems and automatic remote-meter-reading applications, and wireless sensor nets also hold promise for the homeland security market. The sensor market segment exhibiting the fastest growth is the image sensor segment, which is being fueled by breakthroughs in affordability, image resolution, and low power dissipation. Sensor nets’ potential will undoubtedly expand as sensors, transceivers, antennas, batteries, controllers, and communication protocols and topologies continue to improve. It is expected that dust-sized sensors will eventually become de rigueur.
- Typical Applications:
- Automotive Industry.
- Homeland Security.
- Resources:
- Faculty advisor(s): Shawki Areibi
- Students: