LEARNING DATA GENERATION PROGRAM, LEARNING DATA GENERATION APPARATUS AND LEARNING DATA GENERATION METHOD

To provide a learning data generation program, a learning data generation apparatus and a learning data generation method capable of accurately extracting time series images.SOLUTION: A learning data generation program acquires a plurality of images including a predetermined correspondence relations...

Full description

Saved in:
Bibliographic Details
Main Authors ZHENG MINGBIAN, BABA TAKAYUKI, MIYAZAKI NOBUHIRO, MORIWAKI YASUTAKA, SUGIMURA MASAHIKO, TAKEBE HIROAKI, ISHIHARA MASAKI
Format Patent
LanguageEnglish
Japanese
Published 28.05.2020
Subjects
CALCULATING
COMPUTING
COUNTING
IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
PHYSICS
Online AccessGet full text

Cover

More Information
Summary:To provide a learning data generation program, a learning data generation apparatus and a learning data generation method capable of accurately extracting time series images.SOLUTION: A learning data generation program acquires a plurality of images including a predetermined correspondence relationship from a document, determines whether or not coincidence degree of features in the plurality of acquired images meets a predetermined condition, identifies that the plurality of acquired images are time series images when the predetermined condition is determined to be met, and generates learning data including the identified time series data.SELECTED DRAWING: Figure 8 【課題】時系列画像を精度良く抽出することを可能とする学習データ生成プログラム、学習データ生成装置及び学習データ生成方法を提供する。【解決手段】所定の対応関係を有する複数の画像を文献から取得し、取得した複数の画像における特徴量の一致度合が所定の条件を満たすか否かを判定し、所定の条件を満たすと判定した場合、取得した複数の画像が時系列画像であると特定し、特定した時系列画像を含む学習データを生成する。【選択図】図８
Bibliography:Application Number: JP20180212093