A recent article by Ana I. de Castro et al. entitled “3-D Characterization of Vineyards Using a Novel UAV Imagery-Based OBIA Procedure for Precision Viticulture Applications” published in the April 2018 edition of the MDPI Open Access Journal on remote sensing made its way across my desk. In past, I have worked with a number of wine growers regarding how they can best integrate Trimble eCognition image analysis tools into their workflows.
The opening paragraphs of this paper excellently explain the value of remote sensing data for precision viticulture (VP) applications – “to design site-specific management strategies, georeferenced information of the grapevine canopy structure and its variability at the ﬁeld scale are required as input data, since plant architecture is one of the most important traits for the characterization and monitoring of fruit trees”.
eCognition software provides users with the tools to extract such information from their data, in this case very high resolution (VHR) RGB UAV data.
This study investigates an eCognition OBIA-based approach designed to operate without user intervention ( i.e. an automated analysis) with the specific objective:
- Automated classification of grapevines and row gaps
- Automatic estimation of individual grapevines position and size
For this analysis, the authors combined the UAV imagery with a DSM derived from their images in eCognition. The overall accuracy generated by the classification method varied slightly based on the vineyard location, in values were higher than 93.6% in all classifications.
For details on the methodology used, I encourage you to read through this paper. It demonstrates a great use of automated feature extraction from UAV data as well as the data fusion strengths of the eCognition software.