• Design and Development of ICS (Inventory Control System). The system maintained the inventory status, sales orders, purchase orders, sales of each product and customer information. Security of the system was maintained by having privileges and validation was done at all levels, software used was C/JAVA and ORACLE 7.0.

• Developed a computerised solution for organising customer booking for the tyres & tubes for Mypol Sales dept. This project replaced the paper-based system with a computerised booking/invoicing/receipting system. Front end was using Borland Delphi 4.0 and can be connected to any database.

• Designed and Implemented authentication product in Visual Basic 5.0 . This project was to have an authentication system for employers logging into system adding new employers into main database using ODBC.

• Developed a computerised solution for organising customer booking for the tyres & tubes for Sales dept. This project replaced the paper-based system with a computerised booking/invoicing/receipting system. The project involved the design of the Database, forms, and reports using Borland Delphi 4.0.

• Design and Development of employers search engine, the project involved making a employers database using Micorsoft Access and connecting it to application using JDBC & JAVA.

• Patient monitoring system Design and implementation of a Hospital database system for handling patient information data , the diagnosis, the treatment and the patient related medical images. The implementation was done using Java using RMI and JDBC. The database was based on ORACLE.

• Volunteer database for Red-Cross QLD Australia. This project involved the development of a database for Red-Cross using Case Tools (LBMS). The database was implemented using access and the front-end tool was VisualBasic .

• Implemented a Flight Reservation System using ORACLE RDBMS & SQL (Course Work).

• Designed and developed a University Timetabling System the software used was Borland Delphi 4.0 and Paradox database.

• Fast image feature extraction & segmentation included Image feature extraction, Image segmentation and Image retrieval using wavelet transform Using sequential method . This project algorithm could be extended for bar code matching , pattern recognition and character recognition , Implementation involves programming various algorithms in C & UNIX.

• Development of Memory Module and Process Module of operating system using C & UNIX.

• Usage of Socket Programming to Inteconnect two different processes simulated On different machines using distributed systems Enviornment using C & Unix.

This project is based on the utility of Wavelet transform, fractional discrimination and Hausdorff Distance Transform. These features will be used to address some of the questions of this project.

 This project is being supervised by Dr. Jane You and Dr. Hong Shen, School of Computing & Information Technology, Griffith University, Australia.

The objective of this project :

• Image feature extraction
• Image segmentation
• Model-based Matching and Image retrieval
• Compare the computational speed in parallel environment with respect to sequential method.

There are some important differences between the Fourier analysis and wavelets. Fourier basis functions are localized in frequency but not in time, but wavelet is localized in both the domains.

The features of wavelet transform has been applied in this project to extract some feature from the image. The image has been passed through three Wavelet transform. From the figure 1 we can see that one quater of the image is smaller version of the image - a lower resolution version. The other looks like edges of the image. The above hierarchial representation signifies that the image has been coressed to a significant level and the resolution has dropped.

To enhance our feature extraction procedure we have used the Interesting Point Algorithm to determine the edges of the image. The algorithm was first applied to the decomposed original image, and later it was applied to each of the three associated images of subband level three. The sum of these images were used to compare the results with that of the edge detected by the original image. The same has been illustrated in figure 2.

In the project we are presently working on the code for matching and other associated feature extraction algorithms.

• Geoff Davis Wavelet Construction Kit
  
• Embedded predictive wavelet image coder (EPWIC)
  
• WWW - Edge -- Online edge detection
  
• The Computer Vision Home Page
  
• The Digital Image Processing Demo Page
  
• Amara's Wavelet Page
  
• Links to various Wavelet related sites
  
• Image Communication Lab
  
• Hausdorff Distance Transform Group

 

---

All copyrights reserved by the author.
If you have any suggestions please mail the author - ramabellur@mailcity.com.

Dated: Jan 20, 1999

Author: Rama Bellur