Applicability of time-of-flight-based ground and multispectral aerial imaging for grapevine canopy vigour monitoring under direct root-zone deficit irrigation

ABSTRACT

This study investigates the applicability of ground-based 3D time-of-flight (ToF) imaging and small unmanned aerial system (UAS) integrated multispectral imaging as a rapid grapevine canopy vigour mapping tool for decision support during crop production management. Direct root-zone deficit irrigation was applied to grapevines with continuous and pulse irrigation techniques at 15%, 30%, and 60% rates as that of standard irrigation rate (100%) established by the grower in a commercial production. The control block was irrigated continuously at 100% standard irrigation rate. Field plots were imaged using ground platform integrated with 3D ToF imaging sensor and small UAS-integrated multispectral camera at 128 and 65 days before the harvest to estimate the canopy vigour variability associated with irrigation treatments. Customized as well as standard methods (convex hull and voxel grid) were utilized to extract canopy attributes (e.g. volume) from the 3D ToF imaging sensor data. The multispectral images were processed to extract normalized and green normalized difference vegetation index images. Resulting data were used to estimate canopy area ratio i.e. the ratio between the grapevine canopy area with respect to the total area in selected region of interest. A significant correlation (r = 0.34) between canopy volume estimated by customized algorithm and aerial canopy area ratio was observed. Custom canopy volume estimates were also highly correlated (r = 0.79) with voxel grid derived canopy volume data. Overall, 60% continuous direct root zone deficit irrigation appeared to produce canopy volume/vigour comparable with those under control treatment. Results also suggest that grower can utilize either or both (ground and aerial) grapevine canopy mapping techniques for effective management.

Acknowledgements

This project was partially funded by an Internal Competitive Grant on Emerging Research Issues from the Washington State University, and the Washington State Department of Agriculture (WSDA) – Specialty Crop Block Grant Program. Authors would like to acknowledge the suggestions given by Dr. Jianfeng Zhou and Sanaz Jarolmasjed from the Department of Biological Systems Engineering, Washington State University, Pullman, USA.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Washington State Department of Agriculture (WSDA) –Specialty Crop Block Grant Program.