Automated Digital Photo Classification by Tessellated Unit Block Alignment
Chuljin Jang, Hwan-Gue Cho · 2008
Since digital cameras are getting so popular nowadays,a number of digital photos are newly generated at every use of camera. To manage those photos efficiently it is important to reveal how much each photo pair is similar. In this paper we tessellate image into unit blocks and conduct 2D alignment extending to content-based similar region by using seed block pair ranked as high similarity. Through the method, we can distinguish whether photos are sharing the same object or background. Through an alignment, we can get a block region scoring best matching value on whole image. The result is less sensitive to transition or pause change of objects. In experiment, we show how our alignment method is applied to real photo and necessities for further research like photo clustering and massive photo management.