Estimating Rice Agronomic Traits Using Drone-Collected Multispectral Imagery

Abstract

The knowledge of rice nitrogen (N) requirements and uptake capacity are fundamental for the development of improved N management. This paper presents empirical models for predicting agronomic traits that are relevant to yield and N requirements of rice (Oryza sativa L.) through remotely sensed data. Multiple linear regression models were constructed at key growth stages (at tillering and at booting), using as input reflectance values and vegetation indices obtained from a compact multispectral sensor (green, red, red-edge, and near-infrared channels) onboard an unmanned aerial vehicle (UAV). The models were constructed using field data and images from two consecutive years in a number of experimental rice plots in Greece (Thessaloniki Regional Unit), by applying four different N treatments (C0: 0 N kg∙ha−1, C1: 80 N kg∙ha−1, C2: 160 N kg∙ha−1, and C4: 320 N kg∙ha−1). Models for estimating the current crop status (e.g., N uptake at the time of image acquisition) and predicting the future one (e.g., N uptake of grains at maturity) were developed and evaluated. At the tillering stage, high accuracies (R2 ≥ 0.8) were achieved for N uptake and biomass. At the booting stage, similarly high accuracies were achieved for yield, N concentration, N uptake, biomass, and plant height, using inputs from either two or three images. The results of the present study can be useful for providing N recommendations for the two top-dressing fertilizations in rice cultivation, through a cost-efficient workflow.

A Novel Compost for Rice Cultivation Developed by Rice Industrial By-Products to Serve Circular Economy

“…The harvest index (HI) was also calculated as the ratio of grain yield to total dry aboveground biomass. HI is considered as a measure of biological success in partitioning assimilated photosynthate to the harvestable product [21].…”

Modeling Mid-Season Rice Nitrogen Uptake Using Multispectral Satellite Data

“…Site-specific management of rice N application has been shown to both improve yield [19] and reduce total N applied [20]. This improves farming profitability (due to reduced input costs and increased revenue) and environmental outcomes (due to reduced N loss to the atmosphere and water systems) [21]. N requirements also vary spatially within paddies [22], which motivates the application of spatially varying N at PI to match requirements.…”

“…Sensor types include optical cameras (3 bands-red, green blue) [27], multispectral sensors (often adding near-infrared and red-edge bands to optical bands) [21] and hyperspectral sensors [28], which capture reflectance at over many narrow bands. Hyperspectral data was used to investigate determining N content at the heading stage [29], and at the panicle formation stage [28,30].…”

“…Some studies have used hand-held or proximal sensors [31]. Mounting sensors on unmanned aerial vehicles (UAVs) allows collecting within-field variability data in much less time than using proximal sensors [21,33,34]. Some proximal sensing systems use active sensors, where the incident radiation is generated by the device.…”

“…These goals may not be achieved if black-box machine learning methods are used, such as support vector machine (SVM) models and neural networks. Similar to [21], we employed multi-variable linear regression using the Lasso regularization technique to select a subset of the most important predictors, while discarding less important predictors.…”