A deep neural network approach towards real-time on-branch fruit recognition for precision horticulture

作者：

Highlights：

• On-branch fruits can be recognized by color images and a deep learning algorithm.

• The method is independent of the natural environment of orchards.

• Significant improvement on robustness of the algorithm by Global Average Pooling.

• Low running time for ensuring real-time applications in precision horticulture.

摘要

•On-branch fruits can be recognized by color images and a deep learning algorithm.•The method is independent of the natural environment of orchards.•Significant improvement on robustness of the algorithm by Global Average Pooling.•Low running time for ensuring real-time applications in precision horticulture.

论文关键词：Precision horticulture,Fruit recognition,Deep CNN,Global average pooling,Classification

论文评审过程：Received 22 February 2020, Revised 22 May 2020, Accepted 22 May 2020, Available online 29 May 2020, Version of Record 5 June 2020.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.113594