Lahore University of Management Sciences

The Symposium on Document Image Analysis at LUMS-SBASSE

March 21, 2012

Share this page on Facebook

The Symposium on Document Image Analysis was a one-day workshop held on March 20, 2012 at LUMS. The event was jointly organised by the Computer Vision Lab, LUMS-School of Science and Engineering (SBASSE); the Image Understanding and Pattern Recognition (IUPR) Research Group at Technical University of Kaiserslautern (TU Kaiserslautern), Germany and the German Research Centre for Artificial Intelligence (DFKI) along with the support of DAAD-German Academic Exchange Service.

The workshop was designed to give students, professionals and Optical Character Recognition (OCR) enthusiasts a round-up for document image processing specifically in the context of Urdu. It was divided in two parts. The first session, before lunch featured talks introducing document image processing to covering new research directions and challenges in the field. Meanwhile, during the second-leg of the workshop, a hands-on session: “Building your own OCR” was held in Lab 1.

The day started with Dr. Faisal Shafait, the Senior Researcher in Multimedia Analysis and Data Mining (MADM) Competence Centre at DFKI, covering a broad area in text image processing while discussing about its future directions, which was very useful for the participants who actively asked engaging questions.

Dr. Sarmad Hussain- the Head of Centre for Language Engineering at Khawarizmi Institute of Computer Science, University of Engineering & Technology (UET), Lahore, also considered as the pioneering authority in Urdu Image processing- presented intricacies of the Urdu font Nastaliq. His take on Nastaliq being ‘Poetry of Characters’ of the Urdu font.   

Dr. Khurram Khurshid, the Assistant Professor at Institute of Space Technology (IST), Islamabad, presented his work on word spotting in digital libraries. Dr. Arif Zaman, Professor at the Department of Computer Science (CS), LUMS-SBASSE, spoke on his respective area of expertise: Adaptive Optical Marking Technique. Whereas, Syed Saqib Bukhari, the Ph.D. candidate in IUPR Group at TU Kaiserslautern, explained generic text-line extraction method for diverse documents.

All participants actively participated in the question/answer session, which was entertained well by each speaker.

The post-lunch optional session was conducted in Lab 1 where participants were given a hands-on training on developing their own OCR in Python.