Klasifikasi Penyakit Padi berdasarkan Citra Daun Menggunakan Model Terlatih Resnet101

Rice is a staple food source for most countries in the world, including Indonesia. The problem of rice disease is a problem that is quite crucial and is experienced by many farmers. Approximately 200,000 - 300,000 tons per year the amount of rice attacked by pests in Indonesia. Considerable losses are caused by late-diagnosed rice plant diseases that reach a severe stage and cause crop failure. The limited number of Agricultural Extension Officers (PPL) and the Lack of information about disease and proper treatment are some of the causes of delays in handling rice diseases. Therefore, with the development of information technology and computers, it is possible to identify diseases by utilizing Artificial Intelligence, one of which is by using recognition methods based on image processing and pattern recognition technology. The purpose of this research is to create a Machine Learning model by applying the model architecture from Resnet101 combined with the model architecture from the author. The model proposed in this study produces an accuracy of 98.68%.

Detection of Rice Diseases by the Fusion of Optimization based K-means Clustering Algorithm and Faster Region based Convolutional Neural Network

One of the most important food crops in the world is rice, which is highly affected by various plant diseases and pests. Even though there are so many methods to address the concern, detection accuracy is a hectic challenge, which needs to be boosted for an enjoyable farming environment. In the present study a rice disease detection technique was implemented by the fusion of Sailfish optimization – K-means (SCM-KM) and the Faster Region Based Convolutional Neural Network (Faster R-CNN) method. For the optimization of the KM clustering method, Sailfish Optimizer was coupled with the Maximum and Minimum distance algorithm, as well as Chaos theory. The 2D Filtering Mask and Weighted Multilevel Median Filter(2DFM-AMMF) were used to eliminate the sounds. With the aid of the Faster 2D-Otsu technique, the target leaf lesion was segmented from the image. The SCM-KM method is used for detection of rice disease. The Rice diseases were characterized and classified by Region Proposal Networks (RPN) and Faster R-CNN method. Comparative analysis of the SCM-KM+ Faster R-CNN method was performed using the metrics sensitivity, accuracy, and specificity. The proposed detection method produced elevated performance over similar bench marking frameworks.

An Ontology-Based Expert System for Rice Disease Identification and Control Recommendation

A great deal of information related to rice cultivation has been published on the web. Conventionally, this information is studied by end-users to identify pests, and to prevent production losses from rice diseases. Despite its benefits, such information has not yet been encoded in a machine-processable form. This research closes the gap by modeling the knowledge-bases using ontologies and semantic technologies. Our modeled ontologies are externalized from existing reliable sources only, and offer axioms that describe abnormal appearances in rice diseases (and insects) and the corresponding controls. In addition, we developed an expert system called RiceMan, based on our ontologies, to support technical and non-technical users for diagnosing problems from observed abnormalities. We also introduce a composition procedure that aggregates users’ observation data with others for realizing spreadable diseases. This procedure, together with ontology reasoning, lies at the heart of our methodology. Finally, we evaluate our methodology practically with four groups of stakeholders in Thailand: senior agronomists, junior agronomists, agricultural students, and ontology specialists. Both ontologies and RiceMan are evaluated to verify their correctness, usefulness, and usability in various aspects. Our experimental results show that ontology reasoning is a promising approach for this domain problem.

Inducible Enrichment of Osa-miR1432 Confers Rice Bacterial Blight Resistance through Suppressing OsCaML2

MicroRNAs (miRNAs) handle immune response to pathogens by adjusting the function of target genes in plants. However, the experimentally documented miRNA/target modules implicated in the interplay between rice and Xanthomonas oryzae pv. oryzae (Xoo) are still in the early stages. Herein, the expression of osa-miR1432 was induced in resistant genotype IRBB5, but not susceptible genotype IR24, under Xoo strain PXO86 attack. Overexpressed osa-miR1432 heightened rice disease resistance to Xoo, indicated by enhancive enrichment of defense marker genes, raised reactive oxygen species (ROS) levels, repressed bacterial growth and shortened leaf lesion length, whilst the disruptive accumulation of osa-miR1432 accelerated rice susceptibility to Xoo infection. Noticeably, OsCaML2 (LOC_Os03g59770) was experimentally confirmed as a target gene of osa-miR1432, and the overexpressing OsCaML2 transgenic plants exhibited compromised resistance to Xoo infestation. Our results indicate that osa-miR1432 and OsCaML2 were differently responsive to Xoo invasion at the transcriptional level and fine-tune rice resistance to Xoo infection, which may be referable in resistance gene discovery and valuable in the pursuit of improving Xoo resistance in rice breeding.

Hyperspectral Imaging Combined With Deep Transfer Learning for Rice Disease Detection

Various rice diseases threaten the growth of rice. It is of great importance to achieve the rapid and accurate detection of rice diseases for precise disease prevention and control. Hyperspectral imaging (HSI) was performed to detect rice leaf diseases in four different varieties of rice. Considering that it costs much time and energy to develop a classifier for each variety of rice, deep transfer learning was firstly introduced to rice disease detection across different rice varieties. Three deep transfer learning methods were adapted for 12 transfer tasks, namely, fine-tuning, deep CORrelation ALignment (CORAL), and deep domain confusion (DDC). A self-designed convolutional neural network (CNN) was set as the basic network of the deep transfer learning methods. Fine-tuning achieved the best transferable performance with an accuracy of over 88% for the test set of the target domain in the majority of transfer tasks. Deep CORAL obtained an accuracy of over 80% in four of all the transfer tasks, which was superior to that of DDC. A multi-task transfer strategy has been explored with good results, indicating the potential of both pair-wise, and multi-task transfers. A saliency map was used for the visualization of the key wavelength range captured by CNN with and without transfer learning. The results indicated that the wavelength range with and without transfer learning was overlapped to some extent. Overall, the results suggested that deep transfer learning methods could perform rice disease detection across different rice varieties. Hyperspectral imaging, in combination with the deep transfer learning method, is a promising possibility for the efficient and cost-saving field detection of rice diseases among different rice varieties.

Genome-Wide Association Study of QTLs Conferring Resistance to Bacterial Leaf Streak in Rice

Bacterial leaf streak (BLS) is a devastating rice disease caused by the bacterial pathogen, Xanthomonas oryzae pv. oryzicola (Xoc), which can result in severe damage to rice production worldwide. Based on a total of 510 rice accessions, trialed in two seasons and using six different multi-locus GWAS methods (mrMLM, ISIS EM-BLASSO, pLARmEB, FASTmrMLM, FASTmrEMMA and pKWmEB), 79 quantitative trait nucleotides (QTNs) reflecting 69 QTLs for BLS resistance were identified (LOD > 3). The QTNs were distributed on all chromosomes, with the most distributed on chromosome 11, followed by chromosomes 1 and 5. Each QTN had an additive effect of 0.20 (cm) and explained, on average, 2.44% of the phenotypic variance, varying from 0.00–0.92 (cm) and from 0.00–9.86%, respectively. Twenty-five QTNs were detected by at least two methods. Among them, qnBLS11.17 was detected by as many as five methods. Most of the QTNs showed a significant interaction with their environment, but no QTNs were detected in both seasons. By defining the QTL range for each QTN according to the LD half-decay distance, a total of 848 candidate genes were found for nine top QTNs. Among them, more than 10% were annotated to be related to biotic stress resistance, and five showed a significant response to Xoc infection. Our results could facilitate the in-depth study and marker-assisted improvement of rice resistance to BLS.