Toward an automated low-cost three-dimensional crop surface monitoring system using oblique stereo imagery from consumer-grade smart cameras

Abstract

Crop surface models (CSMs) representing plant height above ground level are a useful tool for monitoring in-field crop growth variability and enabling precision agriculture applications. A semiautomated system for generating CSMs was implemented. It combines an Android application running on a set of smart cameras for image acquisition and transmission and a set of Python scripts automating the structure-from-motion (SfM) software package Agisoft Photoscan and ArcGIS. Only ground-control-point (GCP) marking was performed manually. This system was set up on a barley field experiment with nine different barley cultivars in the growing period of 2014. Images were acquired three times a day for a period of two months. CSMs were successfully generated for 95 out of 98 acquisitions between May 2 and June 30. The best linear regressions of the CSM-derived plot-wise averaged plant-heights compared to manual plant height measurements taken at four dates resulted in a coefficient of determination R 2 of 0.87 and a root-mean-square error (RMSE) of 0.08 m, with Willmott's refined index of model performance d r equaling 0.78. In total, 103 mean plot heights were used in the regression based on the noon acquisition time. The presented system succeeded in semiautomatedly monitoring crop height on a plot scale to field scale. © The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

Estimating Barley Biomass with Crop Surface Models from Oblique RGB Imagery

“…Table 2) is comparable to the RMSE of derived plant height based on UAV or LiDAR measurements for other plants (e.g., 3.5 cm for LiDAR measurements on wheat [17], 10-13 cm for UAV-derived plant heights on poppy [18], and 10 cm for winter barley and wheat in [27]) and consistent with the measurements based on the 2014 field experiment presented in [16]. The RMSE of the calibration of the CSM-derived plant height vs. dry biomass regression is consistent if slightly higher when compared to the values achieved in [17] (112 g/m 2 for LiDAR derived plant heights, and 115 g/m 2 for SfM-derived plant heights).…”

“…The general goal of this study was to evaluate if the monitoring system presented by Brocks et al [16] can be used as a robust and cost-effective means for estimating dry and fresh barley biomass on a per-plot scale. Towards this end, CSMs were generated in a semi-automated manner at daily time steps from oblique stereo RGB imagery using a combination of commercial SfM software and custom scripting.…”

“…They can be created from different types of sensors such as LiDAR or RGB cameras using SfM [22] and Multi-View-Stereo (MVS) [23] techniques. In this study, the crop monitoring system presented in [16] was used to generate three CSMs per day during the growing seasons of 2014 and 2015. It uses a pair of smart cameras mounted in an elevated position to acquire oblique RGB imagery of the observed field.…”

“…The RGB images are processed into 2.5D CSMs that represent the canopy surface using a combination of the commercial SfM/MVS software package Agisoft Photoscan (St. Petersburg, Russia) version 1.2.6 and the GIS software package Esri ArcGIS (Redlands, CA, USA) version 10.5. The quality setting "high" was used for the generation of the point clouds, as that resulted in best results in [16]. The points clouds were interpolated to a raster surface using the Inverse Distance Weighted (IDW) algorithm with a target cell size of 1 cm.…”

“…Hence, the installation of a cost-effective continuously operating monitoring station has the advantage of delivering daily imagery to derive certain crop traits. For this purpose, Brocks et al [16] introduced a static low-cost 3D crop-monitoring system which acquires oblique stereo imagery from consumer grade smart cameras for deriving the crop trait plant height. As the authors proved, this system enables the non-destructive determination of plant height in different phenological stages, and the monitoring quality almost equals UAV-based approaches.…”