Creation of effective thumbnails based on handwritten notes

researcher's name
research field
Multimedia database,Computer system,Life/Health/Medical informatics,Software,Intelligent informatics


Tablet terminals which allow handwritten inputs are spreading rapidly, and the amount of handwritten data online has been growing. In response to this, technology is necessary for gathering information effectively from handwritten data. However, traditional thumbnails based on display reduction result in reduced readability of text information handwritten in searches and a corresponding lack of efficiency.


When displaying lists, thumbnails are effective as details can be ascertained easily from an overview, but with handwritten data, the characters can become too small, resulting instead in an increase in time required.
Therefore, reduced thumbnails are produced automatically using snippets extracted of only the important parts, by automatic recognition of expressions emphasized by underlining or capturing in a box, which are enlarged and matched with a diagram. The result allows easy comprehension of the gist of a thumbnail and an increase in search efficiency.


By collaborating with online handwritten data collection systems, such as iPad applications like “Daigaku Note” and Windows applications (input by liquid crystal tablet), efficient handwritten data can be gathered while data can be easily utilized.


Compared with conventional reduced thumbnails, search time can be decreased by 21% and work in terms of number of clicks when searching can be decreased by 51%.

purpose of providing seeds

Sponsord research, Collaboration research, Technical consultation


posted: 2014/05/21