Farm Detection based on Deep Convolutional Neural Nets and Semi-supervised Green Texture Detection using VIS-NIR Satellite Image

Sara Sharifzadeh, Jagati Tata, Bo Tan

Research output: Chapter in Book/Report/Conference proceedingConference proceeding

Abstract

Farm detection using low resolution satellite images is an important topic in digital agriculture. However, it has not received enough attention compared to high-resolution images. Although high resolution images are more efficient for detection of land cover components, the analysis of low-resolution images are yet important due to the low-resolution repositories of the past satellite images used for timeseries analysis, free availability and economic concerns. The current paper addresses the problem of farm detection using low resolution satellite images. In digital agriculture, farm detection has significant role for key applications such as crop yield monitoring. Two main categories of object detection strategies are studied and compared in this paper; First, a two-step semi-supervised methodology is developed using traditional manual feature extraction and modelling techniques; the developed methodology uses the Normalized Difference Moisture Index (NDMI), Grey Level Co-occurrence Matrix (GLCM), 2-D Discrete Cosine Transform (DCT) and morphological features and Support Vector Machine (SVM) for classifier modelling. In the second strategy, high-level features learnt from the massive filter banks of deep Convolutional Neural Networks (CNNs) are utilised. Transfer learning strategies are employed for pretrained Visual Geometry Group Network (VGG-16) networks. Results show the superiority of the high-level features for classification of farm regions.
Original languageEnglish
Title of host publicationDATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications
EditorsSlimane Hammoudi, Christoph Quix, Jorge Bernardino
PublisherSciTePress
Pages100-108
Number of pages9
Volume1
ISBN (Electronic)9789897583773
ISBN (Print)978-989-758-377-3
DOIs
Publication statusPublished - 2019
Event8th International Conference on Data Science, Technology and Applications - VI ENNA HOUSE DIPLOMAT PRAGUE, Prague, Czech Republic
Duration: 26 Jul 201928 Jul 2019
Conference number: 8
http://www.dataconference.org/?y=2019

Publication series

NameDATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications

Conference

Conference8th International Conference on Data Science, Technology and Applications
Abbreviated titleData2019
CountryCzech Republic
CityPrague
Period26/07/1928/07/19
Internet address

Fingerprint

Farms
Image resolution
Textures
Satellites
Neural networks
Agriculture
Discrete cosine transforms
Filter banks
Crops
Support vector machines
Feature extraction
Classifiers
Moisture
Availability
Economics
Geometry
Monitoring

Keywords

  • Classification
  • Convolutional Neural Nets (CNNs)
  • Digital Agriculture
  • Satellite Image
  • Supervised Feature Extraction

ASJC Scopus subject areas

  • Hardware and Architecture
  • Information Systems
  • Software
  • Computer Networks and Communications

Cite this

Sharifzadeh, S., Tata, J., & Tan, B. (2019). Farm Detection based on Deep Convolutional Neural Nets and Semi-supervised Green Texture Detection using VIS-NIR Satellite Image. In S. Hammoudi, C. Quix, & J. Bernardino (Eds.), DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications (Vol. 1, pp. 100-108). (DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications). SciTePress. https://doi.org/10.5220/0007954901000108

Farm Detection based on Deep Convolutional Neural Nets and Semi-supervised Green Texture Detection using VIS-NIR Satellite Image. / Sharifzadeh, Sara; Tata, Jagati; Tan, Bo.

DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications. ed. / Slimane Hammoudi; Christoph Quix; Jorge Bernardino. Vol. 1 SciTePress, 2019. p. 100-108 (DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications).

Research output: Chapter in Book/Report/Conference proceedingConference proceeding

Sharifzadeh, S, Tata, J & Tan, B 2019, Farm Detection based on Deep Convolutional Neural Nets and Semi-supervised Green Texture Detection using VIS-NIR Satellite Image. in S Hammoudi, C Quix & J Bernardino (eds), DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications. vol. 1, DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications, SciTePress, pp. 100-108, 8th International Conference on Data Science, Technology and Applications, Prague, Czech Republic, 26/07/19. https://doi.org/10.5220/0007954901000108
Sharifzadeh S, Tata J, Tan B. Farm Detection based on Deep Convolutional Neural Nets and Semi-supervised Green Texture Detection using VIS-NIR Satellite Image. In Hammoudi S, Quix C, Bernardino J, editors, DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications. Vol. 1. SciTePress. 2019. p. 100-108. (DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications). https://doi.org/10.5220/0007954901000108
Sharifzadeh, Sara ; Tata, Jagati ; Tan, Bo. / Farm Detection based on Deep Convolutional Neural Nets and Semi-supervised Green Texture Detection using VIS-NIR Satellite Image. DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications. editor / Slimane Hammoudi ; Christoph Quix ; Jorge Bernardino. Vol. 1 SciTePress, 2019. pp. 100-108 (DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications).
@inproceedings{19c7a16c893543c2a98dc12a060ddbd3,
title = "Farm Detection based on Deep Convolutional Neural Nets and Semi-supervised Green Texture Detection using VIS-NIR Satellite Image",
abstract = "Farm detection using low resolution satellite images is an important topic in digital agriculture. However, it has not received enough attention compared to high-resolution images. Although high resolution images are more efficient for detection of land cover components, the analysis of low-resolution images are yet important due to the low-resolution repositories of the past satellite images used for timeseries analysis, free availability and economic concerns. The current paper addresses the problem of farm detection using low resolution satellite images. In digital agriculture, farm detection has significant role for key applications such as crop yield monitoring. Two main categories of object detection strategies are studied and compared in this paper; First, a two-step semi-supervised methodology is developed using traditional manual feature extraction and modelling techniques; the developed methodology uses the Normalized Difference Moisture Index (NDMI), Grey Level Co-occurrence Matrix (GLCM), 2-D Discrete Cosine Transform (DCT) and morphological features and Support Vector Machine (SVM) for classifier modelling. In the second strategy, high-level features learnt from the massive filter banks of deep Convolutional Neural Networks (CNNs) are utilised. Transfer learning strategies are employed for pretrained Visual Geometry Group Network (VGG-16) networks. Results show the superiority of the high-level features for classification of farm regions.",
keywords = "Classification, Convolutional Neural Nets (CNNs), Digital Agriculture, Satellite Image, Supervised Feature Extraction",
author = "Sara Sharifzadeh and Jagati Tata and Bo Tan",
year = "2019",
doi = "10.5220/0007954901000108",
language = "English",
isbn = "978-989-758-377-3",
volume = "1",
series = "DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications",
publisher = "SciTePress",
pages = "100--108",
editor = "Slimane Hammoudi and Christoph Quix and Jorge Bernardino",
booktitle = "DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications",

}

TY - GEN

T1 - Farm Detection based on Deep Convolutional Neural Nets and Semi-supervised Green Texture Detection using VIS-NIR Satellite Image

AU - Sharifzadeh, Sara

AU - Tata, Jagati

AU - Tan, Bo

PY - 2019

Y1 - 2019

N2 - Farm detection using low resolution satellite images is an important topic in digital agriculture. However, it has not received enough attention compared to high-resolution images. Although high resolution images are more efficient for detection of land cover components, the analysis of low-resolution images are yet important due to the low-resolution repositories of the past satellite images used for timeseries analysis, free availability and economic concerns. The current paper addresses the problem of farm detection using low resolution satellite images. In digital agriculture, farm detection has significant role for key applications such as crop yield monitoring. Two main categories of object detection strategies are studied and compared in this paper; First, a two-step semi-supervised methodology is developed using traditional manual feature extraction and modelling techniques; the developed methodology uses the Normalized Difference Moisture Index (NDMI), Grey Level Co-occurrence Matrix (GLCM), 2-D Discrete Cosine Transform (DCT) and morphological features and Support Vector Machine (SVM) for classifier modelling. In the second strategy, high-level features learnt from the massive filter banks of deep Convolutional Neural Networks (CNNs) are utilised. Transfer learning strategies are employed for pretrained Visual Geometry Group Network (VGG-16) networks. Results show the superiority of the high-level features for classification of farm regions.

AB - Farm detection using low resolution satellite images is an important topic in digital agriculture. However, it has not received enough attention compared to high-resolution images. Although high resolution images are more efficient for detection of land cover components, the analysis of low-resolution images are yet important due to the low-resolution repositories of the past satellite images used for timeseries analysis, free availability and economic concerns. The current paper addresses the problem of farm detection using low resolution satellite images. In digital agriculture, farm detection has significant role for key applications such as crop yield monitoring. Two main categories of object detection strategies are studied and compared in this paper; First, a two-step semi-supervised methodology is developed using traditional manual feature extraction and modelling techniques; the developed methodology uses the Normalized Difference Moisture Index (NDMI), Grey Level Co-occurrence Matrix (GLCM), 2-D Discrete Cosine Transform (DCT) and morphological features and Support Vector Machine (SVM) for classifier modelling. In the second strategy, high-level features learnt from the massive filter banks of deep Convolutional Neural Networks (CNNs) are utilised. Transfer learning strategies are employed for pretrained Visual Geometry Group Network (VGG-16) networks. Results show the superiority of the high-level features for classification of farm regions.

KW - Classification

KW - Convolutional Neural Nets (CNNs)

KW - Digital Agriculture

KW - Satellite Image

KW - Supervised Feature Extraction

UR - http://www.scopus.com/inward/record.url?scp=85072971833&partnerID=8YFLogxK

U2 - 10.5220/0007954901000108

DO - 10.5220/0007954901000108

M3 - Conference proceeding

SN - 978-989-758-377-3

VL - 1

T3 - DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications

SP - 100

EP - 108

BT - DATA 2019 - Proceedings of the 8th International Conference on Data Science, Technology and Applications

A2 - Hammoudi, Slimane

A2 - Quix, Christoph

A2 - Bernardino, Jorge

PB - SciTePress

ER -