Information technology abroad: New 3-D technology development

Professors Anshuman Razdan and John Femiani’s work with 3-D handwriting has been recognized  by Arizona State University’s Office of Knowledge Enterprise and Development (OKED) and Arizona Technology Enterprise (AzTE). OKED and AzTE are recognizing projects that received patents recently.

Razdan, a resident of Ahwatukee, and Femiani, a Gilbert resident, developed a method to convert a two-dimensional image or bitmap of a handwritten manuscript into 3-D data. The data can be used to automatically recognize features of the manuscript, such as characters or words. The method includes steps to convert the two-dimensional image into 3-D volumetric data, filter the 3-D volumetric data, and process the filtered 3-D volumetric data to resolve features of the two-dimensional image.

The method can be used to differentiate between numerous characteristics that define the overall letter forms in handwritten text, manuscripts or signatures.

“The technology behind the patent will allow a mechanism to separate the handwriting on a page from the print material, analogous to separating wheat from the chaff,” explains Razdan, associate professor in the College of Technology and Innovation’s Department of Engineering at ASU’s Polytechnic campus. 

“The ‘lifting’ of the handwritten ‘layer’ leaves the document to be successfully read or scanned by an optical character recognition (OCR) algorithm, which is used to convert them to digital text that can be searchable, for example.”

The technology works irrespective of the language and can be used to triage millions of documents into a prioritized lists, says Razdan.

“We have further advanced the technologies to be able to detect other features on a document such as logos, tables, figures and diagrams (electrical/mechanical), signatures. This is important because now you can use matching algorithms to see if a specific person signed a particular document,” he says. 

This is not the first patent for Razdan. His other patent is for 3-D Face Recognition.