Mobile Image Recognition Improvements Part 2

The intern is to improve the image recognition software of Semacode Corporation that is intended to automatically detect barcodes in images taken using the camera of a mobile phone, such as the iPhone or Blackberry. The improvements will include: : (1) improve the speed of the “λ-Enhancement” contrast algorithm, (2) solve rotational and skew invariance in 1D barcodes, (3) implement colour thresholding for more accurate scans, (4) complete the iteration of curved 2D barcodes, (5) improve the accuracy of 2D barcodes with a large number of modules, (6) finish optimizing the image recognition algorithm on the iPhone platform, (7) add new support for three other 2D barcode types. The intern will use mathematical techniques to accomplish this goal, applying his knowledge and experience gained in his studies. The partner organization will benefit because it will be able to offer new services oriented towards 1D barcodes, and improve its performance in competition with other companies outside of the country. In addition the company will gain valuable benefits from working with academic partners.

Faculty Supervisor:

Dr. Nico Spronk and Dr. Brian Forrest

Student:

Michael Sgambelluri

Partner:

Semacode Corporation

Discipline:

Mathematics

Sector:

Information and communications technologies

University:

University of Waterloo