IMAGE ANALYSIS APPARATUS, MACHINE EVALUATION APPARATUS, AND MACHINE DIAGNOSIS APPARATUS

To calculate an occupation ratio of each element from an image more easily and quantitatively.SOLUTION: An image analysis apparatus according to the present invention has a learned model 4 obtained by machine learning with a teacher data obtained by adding information of a kind of analysis element t...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: NAKAGAWA MASAMITSU, SUZUKI TAKAHITO
Format: Patent
Sprache:eng ; jpn
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:To calculate an occupation ratio of each element from an image more easily and quantitatively.SOLUTION: An image analysis apparatus according to the present invention has a learned model 4 obtained by machine learning with a teacher data obtained by adding information of a kind of analysis element to each of images obtained by equally dividing an image of an analysis target for learning, an image dividing unit 2C for equally dividing an image of an analysis target into a plurality of images, an image processing unit 2D for sequentially inputting each of the divided images to the learned model 4 to acquire classification information regarding a kind of analysis elements corresponding to each of the divided images, an occupation rate calculating unit 2E for, based on the classification information of each of the images acquired by the image processing unit 2D, calculating, from the number of the divided images corresponding to kinds of the analysis elements to be analyzed, a ration of occupation in the image to be analyzed by the analysis element for the above analysis.SELECTED DRAWING: Figure 1 【課題】より簡易且つ定量的に、画像から各要素の占有割合を求める。【解決手段】学習用の解析対象の画像を複数の画像に等分割した各画像それぞれに解析要素の種類の情報を付加した教師データで機械学習して求めた学習済みモデル4と、解析対象の画像を、複数の画像に等分割する画像分割部2Cと、学習済みモデル4に分割した各画像を順次入力して、分割した各画像それぞれについて、画像に対応する解析要素の種類の分類情報を取得する画像処理部2Dと、画像処理部2Dが取得した画像毎の分類情報に基づき、解析する解析要素の種類に対応する上記分割した画像の数から、当該解析する解析要素が、上記解析対象とする画像中で占有する割合を求める占有割合算出部2Eと、を備える。【選択図】図1