Cotton Yield Estimation Based on Vegetation Indices and Texture Features Derived From RGB Image

Abstract

Yield monitoring is an important parameter to evaluate cotton productivity during cotton harvest. Nondestructive and accurate yield monitoring is of great significance to cotton production. Unmanned aerial vehicle (UAV) remote sensing has fast and repetitive acquisition ability. The visible vegetation indices has the advantages of low cost, small amount of calculation and high resolution. The combination of the UAV and visible vegetation indices has been more and more applied to crop yield monitoring. However, there are some shortcomings in estimating cotton yield based on visible vegetation indices only as the similarity between cotton and mulch film makes it difficult to differentiate them and yields may be saturated based on vegetation index estimates near harvest. Texture feature is another important remote sensing information that can provide geometric information of ground objects and enlarge the spatial information identification based on original image brightness. In this study, RGB images of cotton canopy were acquired by UAV carrying RGB sensors before cotton harvest. The visible vegetation indices and texture features were extracted from RGB images for cotton yield monitoring. Feature parameters were selected in different methods after extracting the information. Linear and nonlinear methods were used to build cotton yield monitoring models based on visible vegetation indices, texture features and their combinations. The results show that (1) vegetation indices and texture features extracted from the ultra-high-resolution RGB images obtained by UAVs were significantly correlated with the cotton yield; (2) The best model was that combined with vegetation indices and texture characteristics RF_ELM model, verification set R2 was 0.9109, and RMSE was 0.91277 t.ha−1. rRMSE was 29.34%. In conclusion, the research results prove that UAV carrying RGB sensor has a certain potential in cotton yield monitoring, which can provide theoretical basis and technical support for field cotton production evaluation.

Estimate leaf area index and leaf chlorophyll content in winter-wheat using image texture and vegetation indices derived from multi-temporal RGB images

“…The R 2 of the linear regression model ranged from 0.52 -0.65 and the R 2 of the machine learning model ranged from 0.57 -0.82. This is similar to the findings of Ma et al (2022) who used machine learning techniques and linear regression models to monitor cotton yield by integrating vegetation indices from RGB images. Zhang et al (2021) also found that the machine learning RF method has a more robust capability than MLR in terms of handling different remote sensing indices when estimating crop growth parameters using UAV, which further reduced the RRMSE by 2.74 -5.11% in the estimation of LAI and LDW.…”

“…Satellites are often limited by revisit periods and weather conditions, making it difficult to provide sufficiently effective data at critical stages of crop growth. Multiple sensors, such as multispectral and hyperspectral sensors, can capture a wealth of information, but storing and processing this data is not straightforward and can easily lead to data redundancy and wastage (Kong et al, 2019;Ma et al, 2022;Zhang et al, 2019). The collection of RGB images by UAV is simple, economical and easy to process (Yamaguchi et al, 2021).…”

“…In assessing aboveground biomass (AGB) in winter oilseed rape fields, Liu et al (Liu et al, 2019) found single texture indices less accurate than vegetation indices (VIs). Machine learning algorithms have been demonstrated to possess robust capabilities for handling non-linear and multivariate regression between remote sensing data and vegetation physiological and biochemical parameters (Ma et al, 2022;Zhang et al, 2022c). The RF model performed better than the SVR model in the three machine learning models, while the BP model was the weakest in all three types of data inputs.…”

Non-destructive monitoring of maize LAI by fusing UAV spectral and textural features

“…Machine learning algorithms combined with remote sensing data have been widely used in areas such as crop growth monitoring (Li et al, 2019;Zheng et al, 2019;Zhang et al, 2021), yield estimation (Fu et al, 2020;Garcıá-Martıńez et al, 2020;Ma et al, 2022) and disease identification (Guo et al, 2021;Zhu et al, 2022). This study used four machine learning algorithms, SVM, RF, BPNN and PLSR, to construct LAI monitoring models for different maize varieties.…”

“…The results of this study are similar to those of previous studies. Ma et al (2022) used color indices and texture features from UAV RGB images to accurately estimate cotton yield, with the RF_ELM model based on color indices and texture features having the highest accuracy (R 2 = 0.911). Yang et al (2021) used vegetation indices and texture features to achieve an estimation of LAI for rice at full fertility.…”