PicClick: two Penn graduate students developing text recognition software

With the recent boom in smartphones, individuals are now able to search the web and find information in a matter of moments. As search engines mature, there is a push for greater advancement in the technology. With the creation of their app PicClick, University of Penn graduate students James Hui and Menglong Zhu are making that search even easier.

PicClick is a text recognition software that allows users to simply snap a photo and click on any text contained within it. The selected text is then run through an algorithm that Hui and Zhu have created and a search result is produced.

Don’t confuse this PicClick with the image search tool of the same name that got some buzz when it launched in 2008.

This Penn project, which started in December 2012, also differs from competitors such as Google Goggle or Bing in its ability to recognize text regardless of the contrast, rotation or perspective in which it is captured.

“We wanted to create a more robust application,” Hui said. “We feel as though if you can see it you should be able to extract it.”

Zhu boasted of PicClick’s superiority: “Other offerings can’t process text that is sideways or skewed. But with ours you can do that.”

Hui and Zhu aim to have the PicClick API released by the end of the year and have been working alongside local coupon website SnipSnap to hone the algorithm and service as a whole before releasing it as a stand alone app.

They have also drawn interest from local investors but Hui says the two are not looking for any money at the moment.

“Right now the technology doesn’t need much capital,” Hui said. “We want to be able to carry ourselves for as long as we can.