A Method for Automatic Generation of Pictorial Maps Using SNS
| Main Authors: | Gao, Shanghui, Ushiama, Taketoshi |
|---|---|
| Format: | Proceeding PeerReviewed Book Thesis |
| Bahasa: | eng |
| Terbitan: |
, 2016
|
| Subjects: | |
| Online Access: |
http://kc.umn.ac.id/3848/1/ADADA_2016_paper_1A-4.pdf http://kc.umn.ac.id/3848/ |
Daftar Isi:
- Pictorial map is a type of map that depicts a given territory with an artistic style rather than technical style. In general, pictorial map contains illustration of difference elements including representative icons such as buildings, people, animals and so on. In addition, those elements should be unique or can show characteristics about a certain area. Otherwise, Instagram has more than 400 million monthly active accounts that share more than 80 million pictures daily. This large volume of image contents includes lots of information about culture and local conditions. In this paper, we present a novel approach that generates a pictorial map automatically using images on Instagram. Our method contains two main aspects: 1) Finding representative location, which is based on the clustering of the images with geographical information. 2) Extraction of typical photos of each representative location.