• 한국멀티미디어학회논문지
• /
• 제25권11호
• /
• pp.1653-1671
• /
• 2022

Crop diseases affect crop production, more than 30 billion USD globally. We proposed a classification study of crop species and diseases using deep learning algorithms for corn, cucumber, pepper, and strawberry. Our study has three steps of species classification, disease detection, and disease classification, which is noteworthy for using captured images without additional processes. We designed deep learning approach of deep learning convolutional neural networks based on Mask R-CNN model to classify crop species. Inception and Resnet models were presented for disease detection and classification sequentially. For classification, we trained Mask R-CNN network and achieved loss value of 0.72 for crop species classification and segmentation. For disease detection, InceptionV3 and ResNet101-V2 models were trained for nodes of crop species on 1,500 images of normal and diseased labels, resulting in the accuracies of 0.984, 0.969, 0.956, and 0.962 for corn, cucumber, pepper, and strawberry by InceptionV3 model with higher accuracy and AUC. For disease classification, InceptionV3 and ResNet 101-V2 models were trained for nodes of crop species on 1,500 images of diseased label, resulting in the accuracies of 0.995 and 0.992 for corn and cucumber by ResNet101 with higher accuracy and AUC whereas 0.940 and 0.988 for pepper and strawberry by Inception.

• 한국멀티미디어학회논문지
• /
• 제23권10호
• /
• pp.1250-1257
• /
• 2020

The early detection of diseases is important in agriculture because diseases are major threats of reducing crop yield for farmers. The shape and color of plant leaf are changed differently according to the disease. So we can detect and estimate the disease by inspecting the visual feature in leaf. This study presents a vision-based leaf classification method for detecting the diseases of tomato crop. ResNet-50 model was used to extract the visual feature in leaf and classify the disease of tomato crop, since the model showed the higher accuracy than the other ResNet models with different depths. We propose a new ensemble approach using several DCNN classifiers that have the same structure but have been trained at different ranges in the DCNN layers. Experimental result achieved accuracy of 97.19% for PlantVillage dataset. It validates that the proposed method effectively classify the disease of tomato crop.

• 산업경영시스템학회지
• /
• 제44권3호
• /
• pp.33-38
• /
• 2021

Recently, transfer learning techniques with a base convolutional neural network (CNN) model have widely gained acceptance in early detection and classification of crop diseases to increase agricultural productivity with reducing disease spread. The transfer learning techniques based classifiers generally achieve over 90% of classification accuracy for crop diseases using dataset of crop leaf images (e.g., PlantVillage dataset), but they have ability to classify only the pre-trained diseases. This paper provides with an evaluation scheme on selecting an effective base CNN model for crop disease transfer learning with regard to the accuracy of trained target crops as well as of untrained target crops. First, we present transfer learning models called CDC (crop disease classification) architecture including widely used base (pre-trained) CNN models. We evaluate each performance of seven base CNN models for four untrained crops. The results of performance evaluation show that the DenseNet201 is one of the best base CNN models.

• 한국시뮬레이션학회논문지
• /
• 제29권4호
• /
• pp.1-7
• /
• 2020

농작물 질병에 대한 조기 진단은 질병의 확산을 억제하고 농업 생산성을 증대하는 데에 있어 중요한 역할을 하고 있다. 최근 합성곱신경망(convolutional neural network, CNN)과 같은 딥러닝 기법을 활용하여 농작물 잎사귀 이미지 데이터세트를 분석하여 농작물 질병을 진단하는 다수의 연구가 진행되었다. 이와 같은 연구를 통해 농작물 질병을 90% 이상의 정확도로 분류할 수 있지만, 사전 학습된 농작물 질병 외에는 진단할 수 없다는 한계를 갖는다. 본 연구에서는 미학습 농작물에 대해 효율적으로 질병 여부를 진단하는 모델을 제안한다. 이를 위해, 먼저 VGG16을 활용한 농작물 질병 분류기(CDC)를 구축하고 PlantVillage 데이터세트을 통해 학습하였다. 이어 미학습 농작물의 질병 진단이 가능하도록 수정된 질병 분류기(mCDC)의 구축방안을 제안하였다. 실험을 통해 본 연구에서 제안한 수정된 질병 분류기(mCDC)가 미학습 농작물의 질병진단에 대해 기존 질병 분류기(CDC)보다 높은 성능을 보임을 확인하였다.

• 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.

• 한국작물학회지
• /
• 제46권1호
• /
• pp.53-56
• /
• 2001

Barley Yellow Dwarf Virus (BYDV), an aphid-borne luteovirus, is a major plant pathogenic disease causing a huge economic loss in the grain production of a wide range of Gramineae species throughout the world. It has been recently reported that BYDV also occurred frequently in wheat field of Korea. Here, we performed to develop the detection and classification methods of BYDV strains that were accomplished by reverse transcription-polymerase chain reaction (RT-PCR). Since there are high variations among BYDV strains, three pairs of primers were designed to detect BYDV strains such as PAV (Vic-PAV and CN-PAV) and MAV (primer A) simultaneously, specifically Vic-PAV(primer B), and MAV (primer C) based on the genomic RNA sequences of BYDV strains previously published. The validity of the primers was confirmed using several BYDV strains obtained from CIMMYT. Though three BYDV strains were able to be detected using primer A, PCR products were not distinguished between two PAV strains. It was possible to separate them with a restriction enzyme, EcoRI, whose restriction site was present in the amplified DNA fragment from Vic-PAV, but not from CN-PAV.

• 식물병연구
• /
• 제17권3호
• /
• pp.265-271
• /
• 2011

우리나라에서 화훼류에 7종류의 파이토플라스마병이 발생하였다. 국화의 Ph-ch1과 Ph-ch2, 나리의 Ph-lily, 페튜니아의 petunia flat stem(PFS-K), 포인세티아의 poinsettia branch inducing(PoiBI-K), 스타티스의 statis witches' broom (SWB-K)과 아잘레아의 azalea witches broom(AWB) 등이다. 16S rRNA 유전자 염기서열을 기본으로 화훼류 파이토플라스마를 분류한 결과 우리나라에는 aster yellow(AY), stolbur와 X-disease 순으로 많이 발생하였다. 파이토플라스마의 특징적인 병징 가운데 하나인 대화증상은 단자엽 식물인 나리와 페튜니아, 포인세티아와 같은 쌍자엽식물에서 모두 발생하였다. 또한 대화증상은 stolbur 그룹의 Ph-lily, AY 그룹의 petunia PFS-K와 X-disease의 포인세티아 PoiBI-K에서 모두 나타났다. 이 결과는 16S rRNA 유전자 염기서열에 기초를 둔 파이토플라스마 분류와 증상과는 일관성있게 일치하지 않는다는 것을 알 수 있다. 우리나라 화훼류에서 발생한 7종의 파이토플라스마를 대추나무빗자루, 오동나무빗자루, 묏대추나무빗자루, 뽕나무 오갈 및 모감주나무파이토플라스마 등 5종의 수목 파이 토플라스마와 16S rRNA 유전자의 염기서열을 비교한 결과 88.5-99.9%의 매우 높은 상동성을 나타내었다. 특히 뽕나무오갈병 파이토플라스마는 PoiBI-K를 제외한 6종의 화훼류 파이토플라스마와 96.3-99.9% 가장 높은 상동성을 나타내었다. 이 결과로 우리나라 화훼류에 발생한 파이토플라스마병은 매개충을 통하여 수목으로부터 전염되었을 것으로 추정되었다.

• Journal of Forest and Environmental Science
• /
• 제40권1호
• /
• pp.1-8
• /
• 2024

Recently, intensive research has been conducted to develop innovative methods for diagnosing plant diseases based on hyperspectral technologies. Hyperspectral analysis is a new subject that combines optical spectroscopy and image analysis methods, which makes it possible to simultaneously evaluate both physiological and morphological parameters. Among the physiological and morphological parameters are classifying healthy and diseased plants, assessing the severity of the disease, differentiating the types of pathogens, and identifying the symptoms of biotic stresses at early stages, including during the incubation period, when the symptoms are not visible to the human eye. Plant diseases cause significant economic losses in agriculture around the world as the symptoms of diseases usually appear when the plants are infected severely. Early detection, quantification, and identification of plant diseases are crucial for the targeted application of plant protection measures in crop production. Hence, this can be done by possible applications of hyperspectral sensors and platforms on different scales for disease diagnosis. Further, the main areas of application of hyperspectral sensors in the diagnosis of plant diseases are considered, such as detection, differentiation, and identification of diseases, estimation of disease severity, and phenotyping of disease resistance of genotypes. This review provides a deeper understanding, of basic principles and implementation of hyperspectral sensors that can measure pathogen-induced changes in plant physiology. Hence, it brings together critically assessed reports and evaluations of researchers who have adopted the use of this application. This review concluded with an overview that hyperspectral sensors, as a non-invasive system of measurement can be adopted in early detection, identification, and possible solutions to farmers as it would empower prior intervention to help moderate against decrease in yield and/or total crop loss.

• 농업과학연구
• /
• 제49권1호
• /
• pp.121-128
• /
• 2022

Fusarium wilt disease caused by Fusarium species is a major problem affecting cultivated cucurbit plants worldwide. Fusarium species are well-known soil-borne pathogenic fungi that cause Fusarium wilt disease in several cucurbit plants. In this study, we aimed to identify and classify pathogenic Fusarium species from cultivated Korean cucurbit plants, specifically watermelon and cucumber. Thirty-six Fusarium isolates from different regions of Korea were obtained from the National Institute of Horticulture and Herbal Science Germplasm collection. Each isolate was morphologically and molecularly identified using an internal transcribed spacer of ribosomal DNA, elongation factor-1α, and the beta-tubulin gene marker sequence. Fusarium species that infect the cucurbit plant family could be divided into three groups: Fusarium oxysporum (F. oxysporum), Fusarium solani (F. solani), and Fusarium equiseti (F. equieti). Among the 36 isolates examined, six were non-pathogenic (F. equiseti: 15-127, F. oxysporum: 14-129, 17-557, 17-559, 18-369, F. solani: 12-155), whereas 30 isolates were pathogenic. Five of the F. solani isolates (11-117, 14-130, 17-554, 17-555, 17-556) were found to be highly pathogenic to both watermelon and cucumber plants, posing a great threat to cucurbit production in Korea. The identification of several isolates of F. equiseti and F. oxysporum, which are both highly pathogenic to bottle gourd, may indicate waning resistance to Fusarium species infection.

• 한국차세대컴퓨팅학회논문지
• /
• 제14권6호
• /
• pp.30-43
• /
• 2018

영상의 특징인 색상, 모양, 질감 등을 이용해 영상을 검색하는 연구들은 많이 진행되어 왔다. 또한 작물의 질병 영상과 관련된 연구들도 진행되고 있다. 농업 현장에서 재배되는 작물에 발생한 질병을 확인하는데 도움이 되기 위해 본 논문에서는 시설원예 작물의 질병 영상을 이용한 유사도 기반 작물 질병 검색 시스템을 제안한다. 제안하는 시스템은 단일 Descriptor를 사용하지 않고, 조합 Descriptor를 통해 기존 대비 영상의 유사도 검색 성능을 높였고 유사도 검색 결과를 가독성 높게 사용자에게 제공하기 위해 가중치 기반 산출방법을 적용했다. 본 논문에서는 총 13개의 개별 Descriptor를 이용해 조합을 진행했다. 조합 Descriptor를 이용해 6개 작물의 질병에 대해 유사도 검색을 진행했고 작물별로 평균 accuracy가 높은 조합 Descriptor를 선정해 유사도 검색에 사용했다. 검색된 결과는 병명의 비율을 기반으로 한 산출방법과 가중치를 기반으로 한 산출방법을 사용해 백분율로 나타냈다. 병명의 비율을 기반으로 한 산출방법은 질의 영상과 유사도 검색에 사용되는 영상의 수가 많은 병명이 1순위로 출력되는 문제점이 있다. 이를 해결하기 위해 가중치를 기반으로 한 산출방법을 사용했다. 작물의 병명별 테스트 영상을 두 가지 산출방법에 적용해 검색 성능을 측정했다. 작물의 질병별로 두 가지 산출방법에 대해 검색 성능 값의 평균을 비교한 결과 고추, 사과 작물에서는 병명의 비율을 기반으로 한 산출방법의 성능이 가중치를 기반으로 한 산출방법의 성능보다 평균 약 11.89%의 높은 성능 결과를 보였다. 국화, 딸기, 배, 포도 작물에서는 가중치를 기반으로 한 산출방법이 병명의 비율을 기반으로 한 산출방법의 성능보다 평균 약 20.34%의 높은 성능 결과를 보였다. 또한 본 논문에서 제안하는 시스템의 UI/UX는 실제 사용자의 피드백을 통해 편리하게 구성했다. 시스템의 화면마다 상단에 제목과 설명을 출력했고 사용자가 질병의 정보를 보기 편리하게 화면을 구성했다. 검색된 질병의 정보는 위에서 제안한 산출방법을 토대로 유사한 질병의 영상과 병명을 출력한다. 시스템의 환경은 PC 환경 기반의 웹 브라우저와 모바일 디바이스 환경 기반의 웹 브라우저를 통해 사용할 수 있도록 구현했다.