|
Title:
|
COMMUNITY BASED FEATURE DATABASE CONSTRUCTION FOR MOBILE IMAGE RETRIEVAL |
|
Author(s):
|
Shigeru Takano, Kensuke Baba, Sozo Inoue |
|
ISBN:
|
978-972-8939-46-5 |
|
Editors:
|
Piet Kommers and Pedro Isaías |
|
Year:
|
2011 |
|
Edition:
|
Single |
|
Keywords:
|
Mobile image retrieval, Wavelet transforms, SNS |
|
Type:
|
Poster/Demonstration |
|
First Page:
|
580 |
|
Last Page:
|
582 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
We propose a novel mobile image retrieval system which combines low-level features (statistics) and high-level features (semantics). The proposed system searches for similar images to a query in a database. The query is a photo captured by a mobile device, and sent to our image retrieval system. Our system can provide a similar image and its information through e-mail immediately. In the first quick answer, their similarities are computed only by using low-level features based on the statistics obtained by applying wavelet filters to the images. On the other hand, high-level features are produced by experts that have enough knowledge of the query or the database images. The moment experts describe some semantics relevant to the target image, our system sends the second answer which is the modified similar image and the detailed information. Finally, we aim to improve the accuracy of the quick answer using the feature database updated by combining the low-level and high-level features. |
|
|
|
|
|
|
