Text Detection Methods in the Scene Images: A Comparison

Abstract

In the present era, the detection of text from scene
images has a significant role in image content
evaluation due to its applicability in different
routine life such as automatic data entry,
augmented computer vision, intelligent content
analysis etc. This is assumed to be a tedious task as
it involves lots of complexities such as diversity in
size, text-font style, orientation, and colour of text
and background factors. To overcome these
complexities, a lot of efficient text detection
techniques are employed. This study illustrates the
tabular comparative analysis of conventional text
detection methods based on techniques used and
their performance metrics. Further the
classification of available datasets for text detection
has also done.