110 likes | 192 Views
SnapWorks. Group 11 Krimy Amichandwala Nihar Gadkari Samantha Misra Urja Shah. SnapWorks. SnapWorks is a photo storing website which provides ways to automate the process of arranging photos based on location and time. Motivation & Features Method Algorithm Take a peek.
E N D
SnapWorks Group 11 Krimy Amichandwala Nihar Gadkari Samantha Misra Urja Shah
SnapWorks SnapWorks is a photo storing website which provides ways to automate the process of arranging photos based on location and time. • Motivation & Features • Method • Algorithm • Take a peek
Motivation & Features Photos form an integral part of one’s memories and information sharing User spends more time in arranging the photos based on events and location Features: • Upload photographs • Overlay them on the map • View clusters based on event • View clusters based on location
Method • Overlay of images on a map – Google Maps API • JavaScript =>ASP. Net
Method Time and Location Information • Metadata of digital images – EXIF tags • Opanda – EXIF Editor • MetaData Extractor
Algorithm • Automatic Organization for Digital Photographs with Geographic Coordinates • The algorithm consists of 3 passes: • Linear pass over sequence of photos • Clustering the photos based on location using dynamic programming algorithm for k-segmentation • Second linear pass merging related photos into a segment
Algorithm • S1, S2, S3 denote the segments formed by 1st pass. • C1, C2 denote the location clusters formed by 2nd pass. • The 3rd pass performs the final event segmentation as shown below
Algorithm • Based on the number of photos in each cluster, perform the k-segmentation algorithm on that cluster recursively. • Algorithm will adapt itself to adjust for a high density region like the San Francisco cluster below. • Naming the clusters will be done by using Google Maps API.
Extensions • Naming the events automatically • Sharing and commenting • Panoramic view • Clustering based on people
References • Automatic organization for digital photographs with geographic coordinates, Mor Naaman, Yee Jiun Song, Andreas Paepcke, Hector GarciaMolina, ACM, 2004. • Time as essence for photo browsing through personal digital libraries, A. Graham, H. Garcia-Molina, A. Paepcke, andT. Winograd, ACM/IEEE-CS, 2002. • Mining Sequential Data - in Search of Segmental Structures, Niina Haiminen (Academic Dissertation) – University of Helsinki, Finland, April 2008