Title:
LINE STRUCTURE EXTRACTION DEVICE, METHOD AND PROGRAM, AND LEARNED MODEL
Document Type and Number:
WIPO Patent Application WO/2020/203552
Kind Code:
A1
Abstract:
The purpose of this invention is to provide a line structure extraction device, method and program with which it is possible to detect a line structure in an image and a learned model. A line structure extraction device according to one embodiment of this disclosure is provided with a learning model which has performed learning to output, as prediction results, one or more component points which constitute a line structure from an image in response to the reception of image input, wherein the learning model contains: a first processing module which receives an image and generates a feature map indicating feature amounts of the image by performing convolution; and a second processing module which calculates, for each unit obtained by dividing the feature map into a plurality of units having regions of a predetermined size in a grid shape, the shift distance from the unit center point to the component point of the line structure closest to the unit center point.
Inventors:
KITAMURA YOSHIRO (JP)
ICHINOSE AKIMICHI (JP)
ICHINOSE AKIMICHI (JP)
Application Number:
PCT/JP2020/013334
Publication Date:
October 08, 2020
Filing Date:
March 25, 2020
Export Citation:
Assignee:
FUJIFILM CORP (JP)
International Classes:
G01T1/161; A61B5/055; A61B6/03; G06T7/00
Foreign References:
| JP2019500110A | 2019-01-10 | |||
| JP2018534694A | 2018-11-22 | |||
| US10198671B1 | 2019-02-05 | |||
| US9858496B2 | 2018-01-02 | |||
| US9858496B2 | 2018-01-02 |
Other References:
REN, SHAOQING ET AL.: "Faster R-CNN: Towards real-time object detection with region proposal networks", ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS, 2015
See also references of EP 3951433A4
See also references of EP 3951433A4
Attorney, Agent or Firm:
MATSUURA, Kenzo (JP)
Download PDF: