Nowadays, many Christian institutions are using digital media to save spiritual pictures, musics or videos, even daily devotional articles which is usually printed monthly. Since many daily devotionals are published in the Internet, it will be difficult to find a daily devotional articles with a spesific topic/category. To make it easier, in this research we use Rocchio’s classification, which use TF-IDF weighting for classification and centroid calculation in every category to classify daily devotional articles. Every testing article will be matched with the centroid using cosine similariy. As a result, the system accuracy is 73,33% using 20% of feature selection. The highest precision goes to Wisdom category which score is 1 for precision by using 100% feature selection. While the highest recall goes to Motivaton category which score is 1 by using 100% feature selection..
Index Terms—classification, categorization, daily devotional article, Rocchio’s classification, centroid, similarity
In this paper, we present a distortion correction for wearable Augmented Reality (AR) on mobile phones. Head Mounted Display (HMD) using mobile phones, such as Samsung Gear VR or Google's cardboard, introduces lens distortion of the rendered image to user. Especially, in case of AR the distortion is more complicated due to the duplicated optical systems from mobile phone's camera and HMD's lens. Furthermore, such distortions generate mismatches of the visual cognition or perception of the user. In a natural way, we can assume that transparent wearable displays are the ultimate visual system which generates the least misperception. Therefore, the image from the mobile phone must be corrected to cancel this distortion to make transparent-like AR display with mobile phone based HMD. We developed a transparent-like display in the mobile wearable AR environment focusing on two issues: pincushion distortion and field-of view. We implemented our technique and evaluated their performance.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.