Deep learning in material recovery: Development of method to create training database

作者：

Highlights：

• Developed method reduces need for material samples.

• Capturing images using multiple illuminations has the biggest impact on performance.

• Material recognition with deep convolutional network reaches human-level accuracy.

摘要

•Developed method reduces need for material samples.•Capturing images using multiple illuminations has the biggest impact on performance.•Material recognition with deep convolutional network reaches human-level accuracy.

论文关键词：Material recognition,Deep neural network,Machine learning,Waste management,Material recovery

论文评审过程：Received 5 September 2018, Revised 25 November 2018, Accepted 30 January 2019, Available online 5 February 2019, Version of Record 12 February 2019.

论文官网地址：https://doi.org/10.1016/j.eswa.2019.01.077