Tools & algorithms to assist in automatically recognizing and deducing information about people in consumer digital images
This project proposes to investigate and test a number of image processing techniques that enhance the identification of persons in consumer digital images. The final product of the research in an integrated software tool that allows for :
- Image marking : load an image from the database and mark important features on it (e.g. faces). Save the location within the picture and the type of marking into the database.
- Algorithm Testing : automatically apply an image processing algorithm (e.g. face detection and recognition) to a set of marked images (only to those that meet a specific criteria)
This masters research in carried out as part of the “Tools and Algorithms to Assist in Automatically Recognizing and Deducing Information about People in Consumer Digital Images” Enterprise Ireland funded project
The Research Team
Dr. Peter Corcoran Mentoring Supervisor NUIG Dr. Dana Vasiloaica Research Supervisor Department of Business & Humanities, IT, Sligo. Mr. Collin Callanan Research Student Department of Business & Humanities, IT Sligo.
Aims and Objectives of the proposed programme of research.
- Research into the following topics
- User Interface Design
- Software Testing
- Distributed Databases
- Prepare a series of test image databases for testing and verification of the algorithms developed as part of the “Tools and Algorithms to Assist in Automatically Recognizing and Deducing Information about People in Consumer Digital Images” Project.
- Develop automatic and semi automatic tools to facilitate the marking and categorisation of the images mentioned in Objective 2.
- Process image data-sets to generate test-sets for specific algorithm verification and comparative evaluation test.
- Develop batch tests to support the testing of algorithms.
- Prepare user manual and help files to assist users in working with the image marking and algorithm testing software application.