A pilot digital image processing approach for detecting vineyard parcels in Douro region through high-resolution aerial imagery

ABSTRACT

Vineyard parcels delimitation is a preliminary but important task to support zoning activities, which can be burdensome and time-consuming when manually performed. In spite of being desirable to overcome such issue, the implementation of a semi-/fully automatic delimitation approach can meet serious development challenges when dealing with vineyards like the ones that prevail in Douro Region (north-east of Portugal), mainly due to the great diversity of parcel/row formats and several factors that can hamper detection as, for example, interrupted rows and inter-row vegetation. Thereby, with the aim of addressing vineyard parcels detection and delimitation in Douro Region, a preliminary method based on segmentation and morphological operations upon highresolution aerial imagery is proposed. This method was tested in a data set collected from vineyards located at the University of Tr'as-os-Montes and Alto Douro(Vila Real, Portugal). The presence of some of the previously mentioned challenging conditions - namely interrupted rows and inter-row grassing - in a few parcels contributed to lower the overall detection accuracy, pointing out the need for future improvements. Notwithstanding, encouraging preliminary results were achieved.

References

1. W Bobillet, JP Da Costa, C Germain, O Lavialle, and G Grenier. Row detection in high resolution remote sensing images of vine fields. Precision Agriculture, page 81, 2003.
2. Lorenzo Comba, Paolo Gay, Jacopo Primicerio, and Davide Ricauda Aimonino. Vineyard detection from unmanned aerial systems images. Computers and Electronics in Agriculture, 114:78--87, 2015.
3. J Primicerio, P Gay, D Ricauda Aimonino, L Comba, A Matese, and SF di Gennaro. Ndvi-based vigour maps production using automatic detection of vine rows in ultra-high resolution aerial images. In Precision agriculture'15, pages 693--712. Wageningen Academic Publishers, 2015.
4. Jacopo Primicerio, Giovanni Caruso, Lorenzo Comba, Alfonso Crisci, Paolo Gay, Silvia Guidoni, Lorenzo Genesio, Davide Ricauda Aimonino, and Francesco Primo Vaccari. Individual plant definitionand missing plant characterization in vineyards from high-resolution uav imagery. European Journal of Remote Sensing, 50(1):179--186, 2017.
5. AP Nolan, S Park, S Fuentes, D Ryu, and H Chung. Automated detection and segmentation of vine rows using high resolution uas imagery in a commercial vineyard. In Proceedings of the 21st International Congress on Modelling and Simulation, Gold Coast, Australia, volume 29, pages 1406--1412, 2015.
6. Nicola Puletti, Rita Perria, and Paolo Storchi. Unsupervised classification of very high remotely sensed images for grapevine rows detection. European Journal of Remote Sensing, 47(1):45--54, 2014.
7. S. Burgos, M. Mota, D. Noll, and B. Cannelle. Use of Very High-Resolution Airborne Images to Analyse 3d Canopy Architecture of a Vineyard. ISPRS -International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, pages 399--403, August 2015.
8. Carlos Poblete-Echeverr'ιa, Guillermo Federico Olmedo, Ben Ingram, and Matthew Bardeen. Detection and segmentation of vine canopy inultra-high spatial resolution rgb imagery obtained from unmanned aerial vehicle (uav): A case study in a commercial vineyard. Remote Sensing, 9(3), 2017.
9. Marie Weiss and Fr'ed'eric Baret. Using 3d point cloudsderived from uav rgb imagery to describe vineyard 3dmacro-structure. Remote Sensing, 9 (2), 2017.
10. Carole Delenne, Sylvie Durrieu, Gilles Rabatel, and Michel Deshayes. From pixel to vine parcel: A complete methodology for vineyard delineation and characterization using remote-sensing data. Computers and Electronics in Agriculture, 70(1):78 -- 83, 2010.
11. Carole Delenne. Extraction et caract'erisation desvignes 'a partir de donn'ees de t'el'ed'etection 'a tr'eshaute r'esolution spatiale: application enLanguedoc-Roussillon pour la constitution de bases dedonn'ees g'eographiques. PhD thesis, Ecole nationale dug'enie rural, des eaux et des for'ets (Paris; Nancy),2006.
12. C Delenne, G Rabatel, V Agurto, and M Deshayes.Vine plot detection in aerial images using fourieranalysis. In 1st International Conference on Object-based Image Analysis (OBIA 2006), 2006.
13. Gilles Rabatel, Carole Delenne, and Michel Deshayes. A non-supervised approach using gabor filters for vine-plot detection in aerial images. Computers and Electronics in Agriculture, 62(2):159 -- 168, 2008.
14. Carole Delenne, Gilles Rabatel, and Michel Deshayes. An automatized frequency analysis for vine plot detection and delineation in remote sensing. IEEE Geoscience and Remote Sensing Letters, 5(3):341--345, 2008.
15. Adam J. Mathews and Jennifer L. R. Jensen. An airborne lidar-based methodology for vineyard parcel detection and delineation. International Journal of Remote Sensing, 33(16):5251--5267, 2012.
16. C. Karakizi and K. Karantzalos. Detecting and classifying vine varieties from very high resolution multispectral data. In 2015 IEEE International Geoscience and Remote Sensing Symposium(IGARSS), pages 3401--3404, July 2015.
17. A. Garc'ιa-Pedrero, C. Gonzalo-Mart'ιn, and M. Lillo-Saavedra. A machine learning approach for agricultural parcel delineation through agglomerative segmentation. International Journal of Remote Sensing, 38(7):1809--1819, 2017.
18. Google Brain. Tensorflow: An open-source software library for machine intelligence. https://www.tensorflow.org/.
19. Fran,cois Chollet et al. Keras. https://github.com/fchollet/keras, 2015.

Index Terms

• Applied computing

  • Computers in other domains

    • Agriculture

• Computing methodologies

  • Computer graphics

    • Image manipulation

      • Image processing

• Information systems

  • Information systems applications

    • Spatial-temporal systems

      • Geographic information systems