Rits Wild Animals Database
www.ritsumei.ac.jp/se/re/izumilab/dist/animals.htm
(C) by Reconfigrable Sysems Lab @ Ritsumeikan U, Mar 2019
Wild Animals in Japan
| Place | Fields and woods in Osaka, Japan |
|---|
| Date | 2016-2018 |
|---|
| Image Size | 32x32 |
|---|
| Format | png |
|---|
| category | #images | examples |
|---|
| Deer | 7262 |
        |
| Boar | 2591 |
        |
| Monkey | 1867 |
        |
| Crow | 161 |
        |
| Bird | 150 |
        |
| Rabbit | 77 |
        |
| Dog | 64 |
        |
| Cat | 55 |
        |
| Fox | 43 |
        |
| Raccoon | 27 |
        |
| Pheasant | 9 |
        |
| Deer_uncertain | 58 |
| Boar_uncertain | 575 |
| Monkey_uncertain | 109 |
| Rabbit_uncertain | 182 |
| Cat_uncertain | 32 |
| Fox_uncertain | 116 |
| Raccoon_uncertain | 40 |
| Pheasant_uncertain | 2 |
| Unidentified | 918 |
- Download from here.
- The original pictures are taken and provided by the Research Institute of Environment, Agriculture and Fisheries, Osaka Prefecture (www.kannousuiken-osaka.or.jp).
- This project is a part of Institute of Disaster Mitigation for Urban Cultural Heritages ( r-dmuch.jp ),
Ritsumeikan University.
- For more details, please refer to ...
-
宮下洸大, 初田慎弥, 孟林, 泉知論:“野生動物自動認識のためのデータベース構築と深層学習の試行”,
画像電子学会 第288回
研究会 講演予稿集, pp.66-71, 2019年3月2日.
- 初田慎弥, 大野真史, 孟林, 泉知論,
"獣害対策のための監視カメラ向けアライグマ検出器の構築と評価",
画像電子学会誌,vol.48,no.2,pp.237-247,May 2019.
Terms of Use
- Users of the database should respect the contributors to the database properly and avoid causing troubles on them.
- Images in the database provided here can be used freely for research and development which need wild animals images as long as the images themselves are not exposed. Statistical results for the images can be disclosed. If the database contributes to the result more than 20 percent in the number of images, the contribution should be acknowledged properly.
- Images themselves can be used only for non-profit academic activities in free-of-charge. Images can be exposed in academic papers (theses, articles, reports, etc.) with adequate acknowledgements. Non-target objects (including humans) accidentally taken in the images should not be exposed.