Optical Character Recognition based Zooming

Abstract

Internet of Things (IoT) is seen as one of the advance paradigm in widespread applications, and also in upcoming generations of information and communication technologies (ICT) with numerous devices from handheld to smart bill board (public displays). The use of expressing one’s view in pictorial way has increased drastically. Currently, there are numerous application where text in video plane plays vital trivial role along with other content. However, sometimes even though size of characters in video plane needs modification in or zoom out), It cannot be done due to lack of technique so. We propose a novel to do (Zoom in mechanism to modify text sizes or Zoom out) in streamed/real time video. There are several zooming techniques available but use of them would potentially affect visual experience (Zoom in would result in loss of content and Zoom outwould in degradation in video quality). Proposed solution uses k-NN algorithm (Optical Character Recognition based algorithm) to find coordinates of area which contains characters in video plane and this coordinates can be used to create texture out of video plane to perform zoom in or out operations.