Title:
IMAGE PROCESSING DEVICE AND PROGRAM
Document Type and Number:
WIPO Patent Application WO/2019/045101
Kind Code:
A1
Abstract:
An image processing device for accepting comics image data and generating, on the basis of the comics image data, information that specifies a frame portion, information that specifies a face portion, information that specifies a body portion, and information that specifies a text portion through use of a result obtained by machine learning so as to specify each of the frame portion, the face position, the body portion, and the text portion, respectively, the information being supplied to prescribed information processing.
Inventors:
AIZAWA KAYOHARU (JP)
OGAWA TORU (JP)
OGAWA TORU (JP)
Application Number:
PCT/JP2018/032635
Publication Date:
March 07, 2019
Filing Date:
September 03, 2018
Export Citation:
Assignee:
UNIV TOKYO (JP)
International Classes:
G06T7/00; G06T7/11
Foreign References:
| JP2011238043A | 2011-11-24 | |||
| JP2002229766A | 2002-08-16 |
Other References:
YANAGISAWA, HIDEAKI ET AL.: "Structural Analysis of Comic Images using Faster R-CNN", PCSJ/IMPS 2016, 30 November 2016 (2016-11-30), pages 80 - 81
FUJIMOTO, AZUMA ET AL.: "Creation and Analysis of Academic Manga Image Dataset with Annotations", IEICE TECHNICAL REPORT, vol. 116, no. 64, 15 March 2017 (2017-03-15), pages 35 - 40
FUKUI, HIROSHI ET AL.: "Research Trends in Pedestrian Detection Using Deep Learning", IEICE TECHNICAL REPORT, vol. 116, no. 366, 17 January 2017 (2017-01-17), pages 37 - 46
FUJIMOTO, AZUMA ET AL.: "Creation and Analysis of Academic Manga Image Dataset with Annotations", IEICE TECHNICAL REPORT, vol. 116, no. 64, 15 March 2017 (2017-03-15), pages 35 - 40
FUKUI, HIROSHI ET AL.: "Research Trends in Pedestrian Detection Using Deep Learning", IEICE TECHNICAL REPORT, vol. 116, no. 366, 17 January 2017 (2017-01-17), pages 37 - 46
Attorney, Agent or Firm:
TAKEI Nobutoshi (JP)
Download PDF: