Painting Alive: Handheld Augmented Reality System for Large Targets Jae-In Hwang, Min-Hyuk Sung, Ig-Jae Kim, Sang Chul Ahn, Hyoung-Gon Kim, and Heedong Ko Imaging Media Research Center, Korea Institute of Science and Technology Abstract. This paper presents a handheld augmented reality (AR) sys- tem and an authoring method which provides alive contents in large targets. In the general augmented reality tools, they are not designed for large targets but for only adequate size of target which fits in the screen. Therefore we designed and built a vision-based AR system and an authoring method that can handle much larger targets than the view frustum. Keywords: augmented reality. 1 Introduction Recently, handheld augmented reality (AR) technology is growing drastically with the advances of mobile devices such as smartphones and tablet-PCs. There are many diversities of application for handheld AR such as games, visual infor- mation providers, advertisements, and so on. One of the adequate applications is mobile tour guide in the museum or sightseeing places [1]. While working on our project named Mobile Augmented Reality Tour (MART) since 2009, we have found several interesting research issues about handheld AR. Because the goal of the project was providing augmented contents through mobile devices during tours in the museum or places, there were many targets that have various sizes and forms. In cases of small targets, we could apply existing augmented reality tracking algorithms or tools such as SIFT [2] or SURF [3]. But most of the track- ing methods were designed for the screen-fit size object. Then what happens if we can see just ten percent of the object through camera in the mobile device? The system would have difficulties in recognizing and tracking the object. So the augmented contents would not appear or could be placed on the wrong position. In the project MART, we had technical challenges to adding augmenting contents on the “Painting of Eastern Palace” which is 576 centimeters in width and 273 centimeters in height. The behavior of tourist using handheld AR tour guide is not predictable. They could focus on any certain part of the painting from various viewing positions and directions. So the technical challenge here was building handheld AR tracking system with the unpredicted view of the large target. In the remaining part of this paper, we will show details of the tracking system. Moreover, we also present about the authoring method for providing various multimedia contents for the objects. R. Shumaker (Ed.): VAMR/HCII 2013, Part I, LNCS 8021, pp. 32–38, 2013. c  Springer-Verlag Berlin Heidelberg 2013