Classification of Paddy Rice Using a Stacked Generalization Approach and the Spectral Mixture Method Based on MODIS Time Series

Abstract

Paddy rice is a major stable food, accounting for about 20% world's food supply. And the rice paddy, an important artificial wetland type, plays an important role in the regional ecological environment. This study proposes a stacked generalization and spectral mixture approach to map paddy rice using coarse spatial resolution images [Moderate Resolution Imaging Spectralradiometer, (MODIS)]. By this method, the time series MODIS enhanced vegetation index images, phenological variables, land surface water index, elevation, and slope images are all employed to produce the optimal feature combination, which is then used to map paddy rice by the stacking algorithm. The validation experiment using the data of the Dongting Lake area showed that the proposed method can improve the overall accuracy of single classifiers, including the support vector machine, random forest, k-nearestneigbor (kNN), extreme gradient boosting (XGB), and decision tree. Stacking (XGB) achieves the highest overall accuracy (90.3%) and Kappa coefficient (0.86), which are 2.8% and 0.03 higher than that of using the single kNN classifier. Furthermore, its user accuracies for distinguishing double-cropping rice and single-season rice are 92.5% and 90.0%, respectively. In terms of the paddy rice classification accuracy, the stacking model is also superior to single classifiers. Moreover, the MODIS-derived rice map obtained by the stacked generalization approach and the spectral mixture method area has a large determination coefficient (R 2 = 0.9975) with the government statistic data. The results demonstrate the potential of the proposed method in using coarse spatial resolution images for large-scale paddy rice mapping. Index Terms-Moderate Resolution Imaging Spectralradiometer (MODIS) time series, paddy rice, phonological variables, spectral mixture, stack generalization. I. INTRODUCTION P ADDY rice feeds 1/5 of the world's people. The demand for paddy rice is increasing due to the fast-growing global population and cause cultivate land and water shortage and biodiversity damage [1]-[5]. As an important cultivated wetland, rice paddy consumes a large amount of fresh water and Manuscript

Rapid nondestructive detection of peanut varieties and peanut mildew based on hyperspectral imaging and stacked machine learning models

“…These are consistent with the findings of several previous studies, e.g. (Zhang et al, 2020) classifies vegetation based on medium resolution spectral imaging technology and SEL, and its accuracy is 5.1-5.2% higher than other single models. (Fu et al, 2022) constructed a model based on multispectral images and SEL, and The training and test set accuracies of the feature wavelengths selected by the four variable selection methods as the model input.…”

“…Stacked ensemble machine learning algorithms have been successfully used in various applications including wind power prediction ( Dong et al., 2021 ), soil classification ( Eyo and Abbey, 2022 ), species classification ( Fu et al., 2022 ), etc. ( Zhang et al., 2020 ) classifies vegetation based on medium resolution spectral imaging technology and SEL, and its accuracy is 5.1-5.2% higher than other single models. ( Fu et al., 2022 ) constructed a model based on multispectral images and SEL, and found that the integrated learning algorithm produced better classification performance than the basic model, with an overall accuracy rate of 1.6-12.7% higher.…”

Identification of Paddy Croplands and Its Stages Using Remote Sensors: A Systematic Review

“…From 2017 there is a spike in the number of published articles, with a drop in 2021. The years 2020 [ 5 , 8 , 9 , 26 , 43 , 47 , 55 , 56 , 57 , 58 , 59 , 60 , 61 , 62 , 63 , 64 , 65 , 66 , 67 , 68 ] and 2022 (until 20 October) [ 2 , 6 , 18 , 28 , 39 , 42 , 44 , 49 , 69 , 70 , 71 , 72 , 73 , 74 , 75 , 76 , 77 , 78 , 79 , 80 ] showed a higher number of articles published, 20 per year, which represents 16% of the total, 32% combined.…”

“…The International Society for Photogrammetry and Remote Sensing (ISPRS) Journal of Photogrammetry and Remote Sensing published six papers [ 9 , 15 , 24 , 86 , 87 , 88 ], representing 6% of the total publications. Institute of Electrical and Electronics Engineers (IEEE) Journal of Selected Topics in Applied earth observation and Remote Sensing (J-STARS) published a total of five papers representing 4% of the total publications [ 17 , 56 , 70 , 89 , 90 ]. Computer and Electronics in Agriculture [ 58 , 79 , 91 , 92 ] and International Journal of Remote Sensing [ 6 , 43 , 93 , 94 ] have published four papers each, representing 3% of the total publications.…”

Machine learning classification of plant genotypes grown under different light conditions through the integration of multi-scale time-series data

“…For example, Random Forest (RF) ML algorithms have been applied to disease prediction, protein sequence selection, and gene selection as well as plant biomass prediction using image-based data [20] , [21] , while Support Vector Machine (SVM) has also been applied to identify plant stress based on image data, neuro-image classification, plant image classification, and biomass prediction [18] , [22] , [23] . As well, stacking multiple ML classifiers has demonstrated additional advantages for crop categorization estimation when compared to the use of a single classifier, suggesting that multiple classifiers in combination can lead to more robust classification outcomes [24] , [25] .…”