Texture defect classification with multiple pooling and filter ensemble based on deep neural network

作者：

Highlights：

• The paper presents a novel Multiple Pooling and Filter approach based on a Deep Neural Network.

• The pre-trained deep network was used to extract features from texture images.

• A consistent and precise detection model is assured for the classification of texture defects.

• The proposed model has high accuracy at or above 95 percent.

摘要

•The paper presents a novel Multiple Pooling and Filter approach based on a Deep Neural Network.•The pre-trained deep network was used to extract features from texture images.•A consistent and precise detection model is assured for the classification of texture defects.•The proposed model has high accuracy at or above 95 percent.

论文关键词：Texture defect recognition,Deep features,Support vector machine,Data augment,Classification

论文评审过程：Received 8 July 2020, Revised 30 December 2020, Accepted 1 March 2021, Available online 10 March 2021, Version of Record 19 March 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.114838