|
Research Automatic text recognition still
remains one of the most challenging problems in computer vision. As text can
be found on any environment, both indoors and outdoors, automatic text
recognition can lead to many different applications, such as support to visually
impaired people, automatic geocoding of businesses, support to robotic
navigation, driver assistance and tourist assistance, among others. My thesis
focuses on text localization and recognition in complex scene images. To
compare my method to the state of the art, I’m working with the datasets used
in both challenges of the Robust Reading Competition held during the ICDAR 2003 and ICDAR 2011
(International Conference on Document Analysis and Recognition). Below you
can see some examples that show the performance of my algorithm in terms of
text localization. I will upload my results and the comparison to other
methods very soon.
I have also
tested my algorithm to the CD Covers
dataset, which is composed of 6,000 CD/DVD cover images. I show some results
below.
|
||||||||||
Home | Research | Publications | Projects | Personal | Links | Contact
|
|||||||||||
