Detail publikačního výsledku

Fruit-CNN: An Efficient Deep learning-based Fruit Classification and Quality Assessment for Precision Agriculture

KUMAR, A.; JOSHI, R.; DUTTA, M.; JONÁK, M.; BURGET, R.

Originální název

Fruit-CNN: An Efficient Deep learning-based Fruit Classification and Quality Assessment for Precision Agriculture

Anglický název

Fruit-CNN: An Efficient Deep learning-based Fruit Classification and Quality Assessment for Precision Agriculture

Druh

Stať ve sborníku v databázi WoS či Scopus

Originální abstrakt

Diseases of fruits and assessment of their quality are one of the key challenges in the farming sector and their automated recognition is very critical to save time and avoid financial loss. The process of manually looking at and identifying the fruit type in crops can be a cumbersome task, the time from which could be put to better use. In this paper, a novel deep learning-based architecture Fruit-CNN has been proposed to identify the type of fruit and their quality assessment of realworld images in multiple visual variations which achieves a test accuracy of 99.6%. The proposed architecture takes minimal time to train the large dataset and test fruit images in comparison with state-of-the-art deep learning models which proves its wide applicability in precision agriculture. More images belonging to various classes can also be trained with fewer parameters which result in fast training of models and less processing time.

Anglický abstrakt

Diseases of fruits and assessment of their quality are one of the key challenges in the farming sector and their automated recognition is very critical to save time and avoid financial loss. The process of manually looking at and identifying the fruit type in crops can be a cumbersome task, the time from which could be put to better use. In this paper, a novel deep learning-based architecture Fruit-CNN has been proposed to identify the type of fruit and their quality assessment of realworld images in multiple visual variations which achieves a test accuracy of 99.6%. The proposed architecture takes minimal time to train the large dataset and test fruit images in comparison with state-of-the-art deep learning models which proves its wide applicability in precision agriculture. More images belonging to various classes can also be trained with fewer parameters which result in fast training of models and less processing time.

Klíčová slova

Deep Learning, Fruits Recognition, Object Identification, Quality assessment, Real-time

Klíčová slova v angličtině

Deep Learning, Fruits Recognition, Object Identification, Quality assessment, Real-time

Autoři

KUMAR, A.; JOSHI, R.; DUTTA, M.; JONÁK, M.; BURGET, R.

Rok RIV

2022

Vydáno

13.12.2021

Nakladatel

IEEE

Místo

Online

ISBN

978-1-6654-0219-4

Kniha

2021 13th International Congress on Ultra Modern Telecommunications and Control Systems and Workshops (ICUMT)

Strany od

60

Strany do

65

Strany počet

6

URL

https://ieeexplore.ieee.org/document/9631643

BibTex

@inproceedings{BUT177009,
  author="Arnav {Kumar} and Rakesh Chandra {Joshi} and Malay Kishore {Dutta} and Martin {Jonák} and Radim {Burget}",
  title="Fruit-CNN: An Efficient Deep learning-based Fruit Classification and Quality Assessment for Precision Agriculture",
  booktitle="2021 13th International Congress on Ultra Modern Telecommunications and Control Systems and Workshops 
(ICUMT)",
  year="2021",
  pages="60--65",
  publisher="IEEE",
  address="Online",
  doi="10.1109/ICUMT54235.2021.9631643",
  isbn="978-1-6654-0219-4",
  url="https://ieeexplore.ieee.org/document/9631643"
}