• Journal of Information Processing Systems
• /
• 제20권2호
• /
• pp.149-158
• /
• 2024

Traditional manual identification of crop leaf diseases is challenging. Owing to the limitations in manpower and resources, it is challenging to explore crop diseases on a large scale. The emergence of artificial intelligence technologies, particularly the extensive application of deep learning technologies, is expected to overcome these challenges and greatly improve the accuracy and efficiency of crop disease identification. Crop leaf disease identification models have been designed and trained using large-scale training data, enabling them to predict different categories of diseases from unlabeled crop leaves. However, these models, which possess strong feature representation capabilities, require substantial training data, and there is often a shortage of such datasets in practical farming scenarios. To address this issue and improve the feature learning abilities of models, this study proposes a deep transfer learning adaptation strategy. The novel proposed method aims to transfer the weights and parameters from pre-trained models in similar large-scale training datasets, such as ImageNet. ImageNet pre-trained weights are adopted and fine-tuned with the features of crop leaf diseases to improve prediction ability. In this study, we collected 16,060 crop leaf disease images, spanning 12 categories, for training. The experimental results demonstrate that an impressive accuracy of 98% is achieved using the proposed method on the transferred ResNet-50 model, thereby confirming the effectiveness of our transfer learning approach.

• The Plant Pathology Journal
• /
• 제39권4호
• /
• pp.319-334
• /
• 2023

Plant disease is an important factor affecting crop yield. With various types and complex conditions, plant diseases cause serious economic losses, as well as modern agriculture constraints. Hence, rapid, accurate, and early identification of crop diseases is of great significance. Recent developments in deep learning, especially convolutional neural network (CNN), have shown impressive performance in plant disease classification. However, most of the existing datasets for plant disease classification are a single background environment rather than a real field environment. In addition, the classification can only obtain the category of a single disease and fail to obtain the location of multiple different diseases, which limits the practical application. Therefore, the object detection method based on CNN can overcome these shortcomings and has broad application prospects. In this study, an annotated apple leaf disease dataset in a real field environment was first constructed to compensate for the lack of existing datasets. Moreover, the Faster R-CNN and YOLOv3 architectures were trained to detect apple leaf diseases in our dataset. Finally, comparative experiments were conducted and a variety of evaluation indicators were analyzed. The experimental results demonstrate that deep learning algorithms represented by YOLOv3 and Faster R-CNN are feasible for plant disease detection and have their own strong points and weaknesses.

• The Plant Pathology Journal
• /
• 제40권2호
• /
• pp.115-124
• /
• 2024

Citrus cultivation plays a pivotal role, making a significant contribution to global fruit production and dietary consumption. Accurate identification of viral pathogens is imperative for the effective management of plant viral disease in citrus crops. High-throughput sequencing serves as an alternative approach, enabling comprehensive pathogen identification on a large scale without requiring pre-existing information. In this study, we employed HTS to investigate viral pathogens infecting citrus in three different regions of South Korea: Jejudo (Jeju), Wando-gun (Wando), and Dangjin-si (Dangjin). The results unveiled diverse viruses and viroids that exhibited regional variations. Notably, alongside the identification of well-known citrus viruses such as satsuma dwarf virus, citrus tatter leaf virus, and citrus leaf blotch virus (CLBV), this study also uncovered several viruses and viroids previously unreported in Korean citrus. Phylogenetic analysis revealed that majority of identified viruses exhibited the closest affilations with isolates from China or Japan. However, CLBV and citrus viroid-I-LSS displayed diverse phylogenetic positions, reflecting their regional origins. This study advances our understanding of citrus virome diversity and regional dynamics through HTS, emphasizing its potential in unraveling intricate viral pathogens in agriculture. Consequently, it significantly contributes to disease management strategies, ensuring the resilience of the citrus industry.

• International Journal of Computer Science & Network Security
• /
• 제23권6호
• /
• pp.115-120
• /
• 2023

Plant leaf diseases and destructive insects are major challenges that affect the agriculture production of the country. Accurate and fast prediction of leaf diseases in crops could help to build-up a suitable treatment technique while considerably reducing the economic and crop losses. In this paper, Convolutional Neural Network based model is proposed to detect leaf diseases of a plant in an efficient manner. Convolutional Neural Network (CNN) is the key technique in Deep learning mainly used for object identification. This model includes an image classifier which is built using machine learning concepts. Tensor Flow runs in the backend and Python programming is used in this model. Previous methods are based on various image processing techniques which are implemented in MATLAB. These methods lack the flexibility of providing good level of accuracy. The proposed system can effectively identify different types of diseases with its ability to deal with complex scenarios from a plant's area. Predictor model is used to precise the disease and showcase the accurate problem which helps in enhancing the noble employment of the farmers. Experimental results indicate that an accuracy of around 93% can be achieved using this model on a prepared Data Set.

• International Journal of Computer Science & Network Security
• /
• 제24권4호
• /
• pp.107-112
• /
• 2024

Plant leaf diseases and destructive insects are major challenges that affect the agriculture production of the country. Accurate and fast prediction of leaf diseases in crops could help to build-up a suitable treatment technique while considerably reducing the economic and crop losses. In this paper, Convolutional Neural Network based model is proposed to detect leaf diseases of a plant in an efficient manner. Convolutional Neural Network (CNN) is the key technique in Deep learning mainly used for object identification. This model includes an image classifier which is built using machine learning concepts. Tensor Flow runs in the backend and Python programming is used in this model. Previous methods are based on various image processing techniques which are implemented in MATLAB. These methods lack the flexibility of providing good level of accuracy. The proposed system can effectively identify different types of diseases with its ability to deal with complex scenarios from a plant's area. Predictor model is used to precise the disease and showcase the accurate problem which helps in enhancing the noble employment of the farmers. Experimental results indicate that an accuracy of around 93% can be achieved using this model on a prepared Data Set.

• 식물병연구
• /
• 제24권4호
• /
• pp.328-331
• /
• 2018

2018년 4월 전북 김제 오이 육묘장에서 오이의 자엽과 본엽에 반점이 형성되고 괴사되는 증상이 관찰되었다. 초기증상은 물방울이 잘 맺히는 자엽 부근에서 시작되었고, 병증상이 확대되면 본엽에 반점이 생기면서 백화 되어 마르는 증상이 관찰되었다. 세균병이 의심되는 식물체를 채집하여 LB agar에 계대하였다. 분리한 한 균주를 가지고 LOPAT test를 수행한 결과, KB agar에서 형광발현, 감자절편에서 무름 증상이 나타나며, Arginine dihydrolase 음성임을 확인하였다. oxidase를 형성하지 않고 담배과민성 실험에서 양성인 것을 확인하여 LOPAT 2그룹에 속하는 것을 확인하였다. 병원성 실험은 발아 후 3주된 오이에 접종하여 접종 3일 후 병징이 동일하게 발생된 것을 확인하였다. 16s rDNA 유전자 염기서열을 이용하여 염기서열분석과 계통수를 분석하여 분리된 균주가 P. viridiflava로 동정되었다. 이 병은 국내에서 처음으로 P. viridiflava에 의해 발생되는 오이점무늬병으로 보고하고자 한다.

• The Plant Pathology Journal
• /
• 제25권1호
• /
• pp.91-94
• /
• 2009

An unreported disease of garlic was observed in commercial fields in Jiangsu province, China. The symptoms started as water soaked lesions at the base of the leaves. Later, water-soaked areas developed on stems and spread to the internal tissues, followed by yellowing and necrosis along leaf edges and soft rot of the stems. The causal organism isolated from symptomatic plants was identified as Pseudomonas fluorescens based on its biochemical and physiological characteristics and confirmed by the cellular fatty acid composition and Biolog data as well as 168 rRNA gene sequence analysis. The bacterial isolates caused similar symptoms when inoculated onto garlic plants. In addition, leek and shallot were susceptible to the P. fluorescens pathogen. However, the P. fluorescens pathogen failed to cause any symptoms when it was inoculated onto 15 other plants. This is the first report of a bacterial disease of garlic caused by P. fluorescens in China.

• The Plant Pathology Journal
• /
• 제39권4호
• /
• pp.409-416
• /
• 2023

Bacterial leaf blight of carrots caused by Xanthomonas hortorum pv. carotae (Xhc) is an important worldwide seed-borne disease. In 2012 and 2013, symptoms similar to bacterial leaf blight were found in carrot farms in Jeju Island, Korea. The phenotypic characteristics of the Korean isolation strains were similar to the type strain of Xhc. Pathogenicity showed symptoms on the 14th day after inoculation on carrot plants. Identification by genetic method was multi-position sequencing of the isolated strain JJ2001 was performed using four genes (danK, gyrB, fyuA, and rpoD). The isolated strain was confirmed to be most similar to Xhc M081. Furthermore, in order to analyze the genetic characteristics of the isolated strain, whole genome analysis was performed through the next-generation sequencing method. The draft genome size of JJ2001 is 5,443,372 bp, which contains 63.57% of G + C and has 4,547 open reading frames. Specifically, the classification of pathovar can be confirmed to be similar to that of the host lineage. Plant pathogenic factors and determinants of the majority of the secretion system are conserved in strain JJ2001. This genetic information enables detailed comparative analysis in the pathovar stage of pathogenic bacteria. Furthermore, these findings provide basic data for the distribution and diagnosis of Xanthomonas hortorum pv. carotae, a major plant pathogen that infects carrots in Korea.

• 한국작물학회지
• /
• 제42권2호
• /
• pp.160-165
• /
• 1997

팥 바이러스병에 대한 기초자료를 얻고자 병징의 유형과 생육 단계별 감량정도 및 수량에 미치는 영향을 조사하고 병징 유형별 바이러스를 분리 동정한 결과는 다음과 같다. 1. 팥에 발생되는 바이러스병의 병징은 크게 mosaic, yellow mosaic 및 severe mosaic의 3가지 군으로 분류되었으며 병징별 분포는 mosaic＞severe mosaic＞yellow mosaic 순이었다. 2. 성숙기까지 매개충 차단 재배시의 이병률은 1.5%(방임구 20.7%)로서 생육 후기에 감량될수록 이병률이 낮았으며 10a당 수량도 171kg으로서 방임구에 비하여 45% 증수되었다. 3. 지표식물 검정 결과 mosaic 유형은 CMV, yellow mosaic 유형은 AMV, severe mosaic 유형은 AzMV의 기주범위와 유사하였다. 4. 항혈청에 의한 각 병징별 반응결과 yellow mosaic 유형의 시요는 AMV 항 혈청과 mosaic 유형의 시요는 CMV의 항혈청과 침강선이 형성되어 각각 AMV 및 CMV의 한 계통으로 판정되었다. 5. 병징 유형별 전자현미경 관찰결과 yellow mosaic유형은 18~58$\times$18nm의 타원형 입자가 다수 관찰되었고 mosaic 증상의 시요에서는 직경 30nm의 구형 입자, severe mosaic병징의 시요에서는 730$\times$12nm의 사상형 입자와 봉입자가 관찰되었다.