A rapid and precise large-scale agricultural disaster survey is a basis for agricultural disaster relief and insurance but is labor-intensive and time-consuming. This study applies Unmanned Aerial Vehicles (UAVs) images through deep-learning image processing to estimate the rice lodging in paddies over a large area. This study establishes an image semantic segmentation model employing two neural network architectures, FCN-AlexNet, and SegNet, whose effects are explored in the interpretation of various object sizes and computation efficiency. Commercial UAVs imaging rice paddies in high-resolution visible images are used to calculate three vegetation indicators to improve the applicability of visible images. The proposed model was trained and tested on a set of UAV images in 2017 and was validated on a set of UAV images in 2019. For the identification of rice lodging on the 2017 UAV images, the F1-score reaches 0.80 and 0.79 for FCN-AlexNet and SegNet, respectively. The F1-score of FCN-AlexNet using RGB + ExGR combination also reaches 0.78 in the 2019 images for validation. The proposed model adopting semantic segmentation networks is proven to have better efficiency, approximately 10 to 15 times faster, and a lower misinterpretation rate than that of the maximum likelihood method.

To meet demand for agriculture products, researchers have recently focused on precision agriculture to increase crop production with less input. Crop detection based on computer vision with unmanned aerial vehicle (UAV)-acquired images plays a vital role in precision agriculture. In recent years, machine learning has been successfully applied in image processing for classification, detection and segmentation. Accordingly, the aim of this study is to detect rice seedlings in paddy fields using transfer learning from two machine learning models, EfficientDet-D0 and Faster R-CNN, and to compare the results to the legacy approach—histograms of oriented gradients (HOG)-based support vector machine (SVM) classification. This study relies on a significant UAV image dataset to build a model to detect tiny rice seedlings. The HOG-SVM classifier was trained and achieved an F1-score of 99% in both training and testing. The performance of HOG-SVM, EfficientDet and Faster R-CNN models, respectively, were measured in mean average precision (mAP), with 70.0%, 95.5% and almost 100% in training and 70.2%, 83.2% and 88.8% in testing, and mean Intersection-over-Union (mIoU), with 46.5%, 67.6% and 99.6% in training and 46.6%, 57.5% and 63.7% in testing. The three models were also measured with three additional datasets acquired on different dates to evaluate model applicability with various imaging conditions. The results demonstrate that both CNN-based models outperform HOG-SVM, with a 10% higher mAP and mIoU. Further, computation speed is at least 1000 times faster than that of HOG-SVM with sliding window. Overall, the adoption of transfer learning allows for rapid establishment of object detection applications with promising performance.

Recently, unmanned aerial vehicles (UAVs) have been broadly applied to the remote sensing field. For a great number of UAV images, deep learning has been reinvigorated and performed many results in agricultural applications. The popular image datasets for deep learning model training are generated for general purpose use, in which the objects, views, and applications are for ordinary scenarios. However, UAV images possess different patterns of images mostly from a look-down perspective. This paper provides a verified annotated dataset of UAV images that are described in data acquisition, data preprocessing, and a showcase of a CNN classification. The dataset collection consists of one multi-rotor UAV platform by flying a planned scouting routine over rice paddies. This paper introduces a semi-auto annotation method with an ExGR index to generate the training data of rice seedlings. For demonstration, this study modified a classical CNN architecture, VGG-16, to run a patch-based rice seedling detection. The k-fold cross-validation was employed to obtain an 80/20 dividing ratio of training/test data. The accuracy of the network increases with the increase of epoch, and all the divisions of the cross-validation dataset achieve a 0.99 accuracy. The rice seedling dataset provides the training-validation dataset, patch-based detection samples, and the ortho-mosaic image of the field.

Rice is one of the three major crops in the world and is the major crop in Asia. Climate change and water resource shortages may result in decreases in rice yields and possible food shortage crises. In this study, water-saving farming management was tested, and IOT field water level monitoring was used to regulate water inflow automatically. Plant height (PH) is an important phenotype to be used to determine difference in rice growth periods and yields using water-saving irrigation. An unmanned aerial vehicle (UAV) with an RGB camera captured sequential images of rice fields to estimate rice PH compared with PH measured on site for estimating rice growth stages. The test results, with two crop harvests in 2019, revealed that with adequate image calibration, the correlation coefficient between UAV-PH and field-PH was higher than 0.98, indicating that UAV images can accurately determine rice PH in the field and rice growth phase. The study demonstrated that water-saving farming is effective, decreasing water usage for the first and second crops of 2019 by 53.5% and 21.7%, respectively, without influencing the growth period and final yield. Coupled with an automated irrigation system, rice farming can be adaptive to water shortage situations.

Grain moisture content (GMC) is a key indicator of the appropriate harvest period of rice. Conventional testing is time-consuming and laborious, thus not to be implemented over vast areas and to enable the estimation of future changes for revealing optimal harvesting. Images of single panicles were shot with smartphones and corrected using a spectral–geometric correction board. In total, 86 panicle samples were obtained each time and then dried at 80 °C for 7 days to acquire the wet-basis GMC. In total, 517 valid samples were obtained, in which 80% was randomly used for training and 20% was used for testing to construct the image-based GMC assessment model. In total, 17 GMC surveys from a total of 201 samples were also performed from an area of 1 m² representing on-site GMC, which enabled a multi-day GMC prediction. Eight color indices were selected using principal component analysis for building four machine learning models, including random forest, multilayer perceptron, support vector regression (SVR), and multivariate linear regression. The SVR model with a MAE of 1.23% was the most suitable for GMC of less than 40%. This study provides a real-time and cost-effective non-destructive GMC measurement using smartphones that enables on-farm prediction of harvest dates and facilitates the harvesting scheduling of agricultural machinery.