Accurate cotton yield prediction is essential for optimizing agricultural practices, improving storage management, and efficiently utilizing resources like fertilizers and water, ultimately benefiting farmers economically. Traditional yield estimation methods, such as field sampling and cotton weighing, are time-consuming and labor intensive. Emerging technologies provide a solution by offering farmers advanced forecasting tools that can significantly enhance production efficiency. In this study, the authors employ segmentation techniques on cotton crops collected using unmanned aerial vehicles (UAVs) to predict yield. The authors apply Segment Anything Model (SAM) for semantic segmentation, combined with You Only Look Once (YOLO) object detection, to enhance the cotton yield prediction model performance. By correlating segmentation outputs with yield data, we implement a linear regression model to predict yield, achieving an R² value of 0.913, indicating the model’s reliability. This approach offers a robust framework for cotton yield prediction, significantly improving accuracy and supporting more informed decision-making in agriculture.

Cotton (Gossypium spp.), a crucial cash crop in the United States, requires the constant monitoring of growth parameters for informed decision-making. Recently, forecasting models have gained prominence for predicting canopy indicators, aiding in-season planning and management decisions to optimize cotton production. This study employed unmanned aerial system (UAS) technology to collect canopy cover (CC) data from a 40-hectare cotton field in Driscoll, Texas, in 2020 and 2021. Long short-term memory (LSTM) models, trained using 2020 data, were subsequently applied to forecast the CC values for 2021. These models were compared with real-time auto-regressive integrated moving average (ARIMA) models to assess their effectiveness in predicting the CC values up to 14 days in advance, starting from the 28th day after crop emergence. The results showed that multiple-input multi-step output LSTM models achieved higher accuracy in predicting the in-season CC values during the early growth stages (up to the 56th day), with an average testing RMSE of 3.86, significantly lower than other single-input LSTM models. Conversely, when sufficient testing data are available, single-input stacked-LSTM models demonstrated precision in CC predictions for later stages, achieving an average RMSE of 3.06. These findings highlight the potential of LSTM models for in-season CC forecasting, facilitating effective management strategies in cotton production.

The rapid decline in water availability for irrigation on the Texas High Plains (THP) is a significant problem affecting crop production and the viability of a large regional economy worth approximately USD 7 billion annually. This region is the largest continuous cotton-producing area in the United States, and the timely delivery and efficient use of irrigation water are critical to the sustainability and profitability of cotton production in this region. Current irrigation scheduling must be improved to reduce water consumption without compromising crop production. Presently, irrigation scheduling based on reference evapotranspiration (ETo) is limited due to the lack of reliable and readily available in-field weather data and updated crop coefficients. Additionally, in-field variability in crop water demand is often overlooked, leading to lower irrigation efficiency. To address these challenges, we explored the potential use of an unmanned aerial vehicle (UAV)-based crop monitoring system to support irrigation management decisions. This study was conducted in Lubbock, Texas, in 2022, where high temporal and spatial resolution images were acquired using a UAV from a cotton field experiment with four irrigation levels. Soil moisture and canopy temperature sensors were deployed to monitor crop response to irrigation and rainfall. The results indicated a significant effect of water stress on crop growth (revealed by UAV-based canopy cover (CC) measurements), yield, and fiber quality. Strong correlations between multi-temporal CC and lint yield (R² = 0.68 to 0.88) emphasized a clear trend: rainfed treatments with lower yields exhibited reduced CC, while irrigated plots with higher CC displayed increased yields. Furthermore, irrigated plots produced more mature and uniform fibers. This study also explored various evapotranspiration calculation approaches indicating that site-specific CC measurements obtained from a UAV could significantly reduce irrigation application. A regression model linking evapotranspiration to canopy cover demonstrated promising potential for estimating water demand in crops with an R² as high as 0.68. The findings highlight the efficacy of UAV-based canopy features in assessing drought effects and managing irrigation water in water-limited production regions like the THP.

Thanks to sensor developments, unmanned aircraft system (UAS) are the most promising modern technologies used to collect imagery datasets that can be utilized to develop agricultural applications in these days. UAS imagery datasets can grow exponentially due to the ultrafine spatial and high temporal resolution capabilities of UAS and sensors. One of the main obstacles to processing UAS data is the intensive computational resource requirements. The structure from motion (SfM) is the most popular algorithm to generate 3D point clouds, orthomosaic images, and digital elevation models (DEMs) in agricultural applications. Recently, the SfM algorithm has been implemented in parallel computing to process big UAS data faster for certain applications. This study evaluated the performance of parallel SfM processing on public cloud computing and on-premise cluster systems. The UAS datasets collected over cropping fields were used for performance evaluation. We used multiple computing nodes and centralized network storage with different network environments for the SfM workflow. In single-node processing, an instance with the most computing power in the cloud computing system performed approximately 20 and 35 percent faster than in the most powerful machine in the on-premises cluster. The parallel processing results showed that the cloud-based system performed better in speed-up and efficiency metrics for scalability, although the absolute processing time was faster in the on-premise cluster. The experimental results also showed that the public cloud computing system could be a good alternative computing environment in UAS data processing for agricultural applications.

Plant growth analysis provides insight regarding the variation behind yield differences in tested genotypes for plant breeders, but adopting this application solely for traditional plant phenotyping remains challenging. Here, we propose a procedure of using uncrewed aerial systems (UAS) to obtain successive phenotype data for growth analysis. The objectives of this study were to obtain high-temporal UAS-based phenotype data for growth analysis and investigate the correlation between the UAS-based phenotype and biomass yield. Seven different energy cane genotypes were grown in a random complete block design with four replications. Twenty-six UAS flight missions were flown throughout the growing season, and canopy cover (CC) and canopy height (CH) measurements were extracted. A five-parameter logistic (5PL) function was fitted through these temporal measurements of CC and CH. The first- and second-order derivatives of this function were calculated to obtain several growth parameters, which were then used to assess the growth of different genotypes with respect to weed competitiveness and biomass yield traits. The results show that CC and CH growth rates significantly differed among genotypes. TH16-16 was outstanding for its ground cover growth; therefore, it was identified as a weed-competitive genotype. Furthermore, TH16-22 had a higher CH maximum growth rate per day, yielding a higher biomass compared to other genotypes. The CH-based multi-temporal data as well as the growth parameters had a better relationship with biomass yield. This study highlights the application of UAS-based high-throughput phenotyping (HTP), along with growth analysis, for assisting plant breeders in decision-making.

Satellite-derived vegetation indices are widely used for crop monitoring, but observations acquired at individual dates may not adequately capture the continuous progression of crop growth and development. This study evaluated a trajectory-based approach for characterizing cotton vegetation dynamics and their relationships with environmental conditions, yield variability, and harvest-aid timing across three dryland growing seasons in South Texas. High-temporal-resolution PlanetScope Normalized Difference Vegetation Index (NDVI) observations were reconstructed as continuous seasonal trajectories using nonlinear curve-fitting functions. Log-normal and asymmetric double-logistic models consistently represented seasonal NDVI trajectories (R² = 0.90–0.98) and enabled extraction of six transition points (T1–T6) representing canopy development, peak canopy condition, and decline, together with rates of change and area-under-the-curve metrics. Transition timing and canopy persistence varied considerably among years under contrasting precipitation and temperature conditions. NDVI and yield correlations strengthened with canopy development and peaked at approximately 95–105 days after planting (r = 0.71–0.80). Among the trajectory-derived parameters, canopy magnitude was the most consistent indicator of within-field yield variability, although the transition contributing most strongly to explained yield variability differed among years. Collectively, 21 vegetation dynamics-based parameters explained 61%, 51%, and 59% of within-field yield variability in 2021, 2022, and 2023, respectively, whereas the contributions of rates of change and cumulative vegetation metrics varied among seasons. Late-season T6 distributions also captured substantial spatial and interannual variation in canopy decline relative to producer-selected harvest-aid timing, although T6 should be considered an indicator of canopy progression rather than physiological boll maturity. Overall, continuous, spatially explicit vegetation trajectories provide a framework for characterizing cotton development, explaining within-field yield variability, and potentially supporting in-season crop management beyond vegetation-index observations acquired at fixed dates.