• Korean Journal of Remote Sensing
• /
• v.40 no.1
• /
• pp.57-69
• /
• 2024

Crop classification plays a vitalrole in monitoring agricultural landscapes and enhancing food production. In this study, we explore the effectiveness of Long Short-Term Memory (LSTM) models for crop classification, focusing on distinguishing between apple and rice crops. The aim wasto overcome the challenges associatedwith finding phenology-based classification thresholds by utilizing LSTM to capture the entire Normalized Difference Vegetation Index (NDVI)trend. Our methodology involvestraining the LSTM model using a reference site and applying it to three separate three test sites. Firstly, we generated 25 NDVI imagesfrom the Sentinel-2A data. Aftersegmenting study areas, we calculated the mean NDVI values for each segment. For the reference area, employed a training approach utilizing the NDVI trend line. This trend line served as the basis for training our crop classification model. Following the training phase, we applied the trained model to three separate test sites. The results demonstrated a high overall accuracy of 0.92 and a kappa coefficient of 0.85 for the reference site. The overall accuracies for the test sites were also favorable, ranging from 0.88 to 0.92, indicating successful classification outcomes. We also found that certain phenological metrics can be less effective in crop classification therefore limitations of relying solely on phenological map thresholds and emphasizes the challenges in detecting phenology in real-time, particularly in the early stages of crops. Our study demonstrates the potential of LSTM models in crop classification tasks, showcasing their ability to capture temporal dependencies and analyze timeseriesremote sensing data.While limitations exist in capturing specific phenological events, the integration of alternative approaches holds promise for enhancing classification accuracy. By leveraging advanced techniques and considering the specific challenges of agricultural landscapes, we can continue to refine crop classification models and support agricultural management practices.

• Proceedings of the Korean Society of Crop Science Conference
• /
• 2022.10a
• /
• pp.80-80
• /
• 2022

Estimation of heading date of paddy rice is laborious and time consuming. Therefore, automatic estimation of heading date of paddy rice is highly essential. In this experiment, deep learning classification models were used to classify two difference categories of rice (vegetative and reproductive stage) based on the panicle initiation of paddy field. Specifically, the dataset includes 444 slanted view images belonging to two categories and was then expanded to include 1,497 images via IMGAUG data augmentation technique. We adopt two transfer learning strategies: (First, used transferring model weights already trained on ImageNet to six classification network models: VGGNet, ResNet, DenseNet, InceptionV3, Xception and MobileNet, Second, fine-tuned some layers of the network according to our dataset). After training the CNN model, we used several evaluation metrics commonly used for classification tasks, including Accuracy, Precision, Recall, and F1-score. In addition, GradCAM was used to generate visual explanations for each image patch. Experimental results showed that the InceptionV3 is the best performing model in terms of the accuracy, average recall, precision, and F1-score. The fine-tuned InceptionV3 model achieved an overall classification accuracy of 0.95 with a high F1-score of 0.95. Our CNN model also represented the change of rice heading date under different date of transplanting. This study demonstrated that image based deep learning model can reliably be used as an automatic monitoring system to detect the heading date of rice crops using CCTV camera.

• 한국약용작물학회:학술대회논문집
• /
• 2010.10a
• /
• pp.139-139
• /
• 2010

• Journal of Climate Change Research
• /
• v.6 no.1
• /
• pp.21-31
• /
• 2015

Various crop models have been extensively used for estimation of the crop yields. Compared to the other models, the EPIC model uses a unified approach to simulate more than 100 types of crops. It has been successfully applied in simulating crop yields for various combinations of weather conditions, soil properties, crops, and management schemes in many countries. The objective of this study was to estimate the rice and maize yield in South Korea using the EPIC model. The input datasets for the 30 types in the 11 categories were created for the EPIC model. The EPIC model simulated rice and maize yields. The performance of the EPIC model was evaluated with the goodness-of-fit measures including Root Mean Square Error (RMSE), Relative Error (RE), Nash-Sutcliffe Efficiency Coefficient (NSEC), Mean Absolute Error (MAE), and Pearson Correelation Coefficient (r). The rice yield showed to more high accuracy than maize yield on four type of method without NSEC. Theses results showed that the EPIC model better simulated rice yields than maize yields. The results suggest that the EPIC crop model can be useful to estimate crop yield in South Korea.

• Proceedings of the Korean Society of Crop Science Conference
• /
• 2017.06a
• /
• pp.14-14
• /
• 2017

The discipline of plant breeding is experiencing a renaissance impacting crop improvement as a result of new technologies, however fundamental questions remain for predicting the phenotype and how the environment and genetics shape it. Inexpensive DNA sequencing, genotyping, new statistical methods, high throughput phenotyping and gene-editing are revolutionizing breeding methods and strategies for improving both quantitative and qualitative traits. Genomic selection (GS) models use genome-wide markers to predict performance for both phenotyped and non-phenotyped individuals. Aerial and ground imaging systems generate data on correlated traits such as canopy temperature and normalized difference vegetative index that can be combined with genotypes in multivariate models to further increase prediction accuracy and reduce the cost of advanced trials with limited replication in time and space. Design of a GS training population is crucial to the accuracy of prediction models and can be affected by many factors including population structure and composition. Prediction models can incorporate performance over multiple environments and assess GxE effects to identify a highly predictive subset of environments. We have developed a methodology for analyzing unbalanced datasets using genome-wide marker effects to group environments and identify outlier environments. Environmental covariates can be identified using a crop model and used in a GS model to predict GxE in unobserved environments and to predict performance in climate change scenarios. These new tools and knowledge challenge the plant breeder to ask the right questions and choose the tools that are appropriate for their crop and target traits. Contemporary plant breeding requires teams of people with expertise in genetics, phenotyping and statistics to improve efficiency and increase prediction accuracy in terms of genotypes, experimental design and environment sampling.

• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.53 no.1
• /
• pp.1-7
• /
• 2008

This experiment was carried out to evaluate the agronomic stability of North Korean rice varieties using the statistical model developed by Grafius, Finlay, and Ever hart. The lowest yearly variation based on coefficients of variation was found in Hannam 29 for number of panicles per hill, in Sijoong 9 for number of grains per panicle, in Pyeongyang 3 for ripened grain ratio, in Sijoong 16 for 1,000 grain weight, and in Yeomju 1 for grain yield. By Grafius's model, Pyeongbook 3, Weonsan 66 in early maturing groups and Seohaechalbyeo in medium maturing groups show stable for 3 years. Weonsan 66 in early maturing groups and Seohaechalbyeo in medium maturing groups were found to be highly stable as analyzed by both Finlay and Wilkinson's model and Everhart & Russell's model. With reference to three models, Weonsan 66 was highly stable for 3 years with showing more yield than Odaebyeo in early maturing groups while Seohaechalbyeo was highly stable for 3 years with showing high yield than Hwaseongbyeo in medium maturing groups above $5\;t\;ha^{-1}$ of milled rice respectively.

• Proceedings of the Korean Society of Crop Science Conference
• /
• 2022.10a
• /
• pp.67-67
• /
• 2022

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.20 no.2
• /
• pp.228-232
• /
• 2018

Process-based crop models have been used to assess the impact of climate change on crop production. These models are implemented in procedural or object oriented computer programming languages including FORTRAN, C++, Delphi, Java, which have a stiff learning curve. The requirement for a high level of computer programming is one of barriers for efforts to develop and improve crop models based on biophysical process. In this study, we attempted to develop a Chinese cabbage model using Microsoft Excel with Visual Basic for Application (VBA), which would be easy enough for most agricultural scientists to develop a simple model for crop growth simulation. Results from Soil-Plant-Atmosphere-Research (SPAR) experiments under six temperature conditions were used to determine parameters of the Chinese cabbage model. During a plant growing season in SPAR chambers, numbers of leaves, leaf areas, growth rate of plants were measured six times. Leaf photosynthesis was also measured using LI-6400 Potable Photosynthesis System. Farquhar, von Caemmerer, and Berry (FvCB) model was used to simulate a leaf-level photosynthesis process. A sun/shade model was used to scale up to canopy-level photosynthesis. An Excel add-in, which is a small VBA program to assist crop modeling, was used to implement a Chinese cabbage model under the environment of Excel organizing all of equations into a single set of crop model. The model was able to simulate hourly changes in photosynthesis, growth rate, and other physiological variables using meteorological input data. Estimates and measurements of dry weight obtained from six SPAR chambers were linearly related ($R^2=0.985$). This result indicated that the Excel/VBA can be widely used for many crop scientists to develop crop models.

• Proceedings of the Korean Society of Crop Science Conference
• /
• 2005.08a
• /
• pp.22-46
• /
• 2005

Spatial and timely information on crop and filed conditions is one of the most important basics for rational and efficient planning and management in agriculture. Remote sensing, GIS, and modeling are powerful tools for such applications. This paper presents an overview of the state of the art in remote sensing of crop and field conditions with some case studies. It is also shown that a synergistic linkage between process-based models and remote sensing signatures enables us to estimate the multiple crop/ecosystem variables at a dynamic mode. Remotely sensed information can greatly reduce the uncertainty of simulation models by compensating for insufficient availability of data or parameters. This synergistic approach allows the effective use of infrequent and multi-source remote sensing data for estimating important ecosystem variables such as biomass growth and ecosystem $CO_2$ flux. This paper also shows a geo-spatial information system that enables us to integrate, search, extract, process, transform, and calculate any part of the data based on ID#, attributes, and/or by river-basin boundary, administrative boundary, or boundaries of arbitrary shape/size all over Japan. A case study using the system demonstrates that the nitrogen load from fertilizer was closely related to nitrate concentration of groundwater. The combined use of remote sensing, GIS and modeling would have great potential for various agro-ecosystem applications.

• Korean Journal of Remote Sensing
• /
• v.38 no.2
• /
• pp.199-213
• /
• 2022

The unsupervised domain adaptation can solve the impractical issue of repeatedly collecting high-quality training data every year for annual crop classification. This study evaluates the applicability of deep learning-based unsupervised domain adaptation models for crop classification. Three unsupervised domain adaptation models including a deep adaptation network (DAN), a deep reconstruction-classification network, and a domain adversarial neural network (DANN) are quantitatively compared via a crop classification experiment using unmanned aerial vehicle images in Hapcheon-gun and Changnyeong-gun, the major garlic and onion cultivation areas in Korea. As source baseline and target baseline models, convolutional neural networks (CNNs) are additionally applied to evaluate the classification performance of the unsupervised domain adaptation models. The three unsupervised domain adaptation models outperformed the source baseline CNN, but the different classification performances were observed depending on the degree of inconsistency between data distributions in source and target images. The classification accuracy of DAN was higher than that of the other two models when the inconsistency between source and target images was low, whereas DANN has the best classification performance when the inconsistency between source and target images was high. Therefore, the extent to which data distributions of the source and target images match should be considered to select the best unsupervised domain adaptation model to generate reliable classification results.