Elshikha, D., Roanhorse, A., Waller, P., Jenkins, V. (2007). REMOTELY PILOTED VEHICLES AND PRECISION AGRICULTURE APPLICATIONS. Journal of Soil Sciences and Agricultural Engineering, 32(1), 503-517. doi: 10.21608/jssae.2007.200946
D. E. Elshikha; A. R. Roanhorse; P. M. Waller; V. Jenkins. "REMOTELY PILOTED VEHICLES AND PRECISION AGRICULTURE APPLICATIONS". Journal of Soil Sciences and Agricultural Engineering, 32, 1, 2007, 503-517. doi: 10.21608/jssae.2007.200946
Elshikha, D., Roanhorse, A., Waller, P., Jenkins, V. (2007). 'REMOTELY PILOTED VEHICLES AND PRECISION AGRICULTURE APPLICATIONS', Journal of Soil Sciences and Agricultural Engineering, 32(1), pp. 503-517. doi: 10.21608/jssae.2007.200946
Elshikha, D., Roanhorse, A., Waller, P., Jenkins, V. REMOTELY PILOTED VEHICLES AND PRECISION AGRICULTURE APPLICATIONS. Journal of Soil Sciences and Agricultural Engineering, 2007; 32(1): 503-517. doi: 10.21608/jssae.2007.200946
REMOTELY PILOTED VEHICLES AND PRECISION AGRICULTURE APPLICATIONS
1Agric. Engineering depart. Univ. of Mansoura, Mansoura, Egypt
2Agric. and Biosystems Engineering Depart. Univ.of Arizona, Tucson, Arizona, USA
3Radio Control Model Aircraft, Inc., Tucson, Arizona, USA
Abstract
Extensive research has been conducted using satellite and aerial imagery to observe vegetation status. Nonetheless, these platforms have limitations, particularly with regard to cost, limited spatial and temporal resolution. Remotely piloted vehicles (RPV) are effective, low cost alternatives to satellites and manned aircraft for acquiring high-resolution images. In this study, remotely piloted vehicles (RPV) constructed from a commercially available remote control model airplane was used as a remote sensing platform to determine normalized difference vegetation index (NDVI) as a measure of percent canopy cover over a controlled experimental sight in central Arizona.
The RPV collected digital images above a controlled experimental site with a multi-spectral camera (DYCAM) with spectral bands at 600 nm and 1100 nm, visible red and near infrared (NIR), respectively. The experimental design included 12 different treatment combinations of nitrogen, water, and plant density. Two levels of nitrogen were applied, which were “high” at 150 kg N ha-1 (63 kg N / Feddan) and “low” at 93 kg N ha-1 (39 kg N / Feddan). Plant density treatments were: sparse (90 plants m-2, single line planting); typical (164 plants m-2, single-line planting) and dense (291 plants m-2, double-line planting). Water treatments included comparison of two different calculation methods for crop coefficient: FAO-based versus remote sensing-based. Three RPV flights at days of year 57, 71 and 113 were the focus of this study. Field measurements of percent canopy cover (canopy width) and hand held radiometer data at these three days constituted ground truthing for the RPV data. NDVI data of the DYCAM and the hand-held radiometer were compared to the field measurements of percent canopy cover. For the first two data days, the NDVI developed with the DYCAM (i.e. RPV) was highly correlated to both percent canopy cover and NDVI developed with the Exotech radiometer (ground data). Though, due to senescence on DOY 113, DYCAM was poorly correlated to Exotech and canopy percent cover. The high correlation of the NDVI to percent cover demonstrates a potential of RPVs as a convenient, cost effective, real-time tool in precision agriculture. The ease in obtaining flying lessons, the accessibility of the platform and sensors, and the associated cost savings may make the RPVs appealing tools for monitoring crop health to agricultural producers and academia as well.
View on SCiNiTO
Top