大学サイト   学部サイト   学科サイト   EnglishPages
 
旧尾家・塚本研究室です.尾家祐二 現・九州工業大学学長の情報は こちら.
   
研究業績


画像圧縮率に基づく画像分類の有効性評価
著者
  宇留島 智紀, 坂戸 響, 大西 圭
 
雑誌名/会議名
  第16回日本知能情報ファジィ学会九州支部学術講演会予稿集, 福岡県北九州市, 2015年3月. 2015 年 3 月
 
アブストラクト
  We previously proposed to use an image compression ratio as a feature of an image. In this study we evaluate the effectiveness of that idea for image classification by subjective tests and statistical hypothesis testing. In the subjective tests, subjects see a group of images with similar compression ratios to a key image and that of randomly chosen images, and then answer which group is more similar to the key image. Then, we apply a sign test to data of the two groups and show that there is a statistically significant difference between the two group. This result indicates that an image compression ratio as one of image features is basically effective for image classification
 
キーワード
 
 
記述言語
  Japanese
 
 

web_admin [ at ] infonet.cse.kyutech.ac.jp /  Copyright (c) 2016 Oie & Tsukamoto Laboratory, All Right Reserved.