Presentation 1998/3/6
Color Image Compression using Neural Network
Katsuhiro FUJIMOTO, Akira INOUE, Shun-ichi IISAKU,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In compressing(encoding )the image data using neural network.there has been a problem that the decoded accuracy was law according to the selection of the pixel blocks which are used to the supervised data.Moreover the evaluation with applying one to color image has not been sufficient yet.In this paper.we propose two methods for this scheme.One is the supervised data selection by clustering the elementary statistics of the pixel blocks to improve the decoded accuracy.another is the network structure for color image to reduce the calculating time. With the result of the experiments. it was confirmed that these methods was effective.In conclusion we also summarize the advantage of image data compression using neural network including a comparison with the JPEG method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) color image compression / image coding / neural network / supervised data selection / clustering
Paper #
Date of Issue

Conference Information
Committee IE
Conference Date 1998/3/6(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Image Engineering (IE)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Color Image Compression using Neural Network
Sub Title (in English)
Keyword(1) color image compression
Keyword(2) image coding
Keyword(3) neural network
Keyword(4) supervised data selection
Keyword(5) clustering
1st Author's Name Katsuhiro FUJIMOTO
1st Author's Affiliation Mken Co., Ltd()
2nd Author's Name Akira INOUE
2nd Author's Affiliation Mken Co., Ltd
3rd Author's Name Shun-ichi IISAKU
3rd Author's Affiliation Communications Research Laboratory, MPT
Date 1998/3/6
Paper #
Volume (vol) vol.97
Number (no) 590
Page pp.pp.-
#Pages 6
Date of Issue