VGG16을 활용한 미학습 농작물의 효율적인 질병 진단 모델

정석봉,윤협상 한국시뮬레이션학회 2020 한국시뮬레이션학회 논문지 Vol.29 No.4

Early detection and classification of crop diseases play significant role to help farmers to reduce disease spread and to increase agricultural productivity. Recently, many researchers have used deep learning techniques like convolutional neural network (CNN) classifier for crop disease inspection with dataset of crop leaf images (e.g., PlantVillage dataset). These researches present over 90% of classification accuracy for crop diseases, but they have ability to detect only the pre-trained diseases. This paper proposes an efficient disease inspection CNN model for new crops not used in the pre-trained model. First, we present a benchmark crop disease classifier (CDC) for the crops in PlantVillage dataset using VGG16. Then we build a modified crop disease classifier (mCDC) to inspect diseases for untrained crops. The performance evaluation results show that the proposed model outperforms the benchmark classifier. 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 질병 진단에 대해 기존 질병 분류기(CDC)보다 높은 성능을 보임을 확인하였다.

Descriptor 조합 및 동일 병명 이미지 수량 역비율 가중치를 적용한 유사도 기반 작물 질병 검색 기술 설계 및 구현

임혜진,정다운,유성준,구영현,박종한 한국차세대컴퓨팅학회 2018 한국차세대컴퓨팅학회 논문지 Vol.14 No.6

Many studies have been carried out to retrieve images using colors, shapes, and textures which are characteristic of images. In addition, there is also progress in research related to the disease images of the crop. In this paper, to be a help to identify the disease occurred in crops grown in the agricultural field, we propose a similarity-based crop disease search system using the diseases image of horticulture crops. The proposed system improves the similarity retrieval performance compared to existing ones through the combination descriptor without using a single descriptor and applied the weight based calculation method to provide users with highly readable similarity search results. In this paper, a total of 13 Descriptors were used in combination. We used to retrieval of disease of six crops using a combination Descriptor, and a combination Descriptor with the highest average accuracy for each crop was selected as a combination Descriptor for the crop. The retrieved result were expressed as a percentage using the calculation method based on the ratio of disease names, and calculation method based on the weight. The calculation method based on the ratio of disease name has a problem in that number of images used in the query image and similarity search was output in a first order. To solve this problem, we used a calculation method based on weight. We applied the test image of each disease name to each of the two calculation methods to measure the classification performance of the retrieval results. We compared averages of retrieval performance for two calculation method for each crop. In cases of red pepper and apple, the performance of the calculation method based on the ratio of disease names was about 11.89% on average higher than that of the calculation method based on weight, respectively. In cases of chrysanthemum, strawberry, pear, and grape, the performance of the calculation method based on the weight was about 20.34% on average higher than that of the calculation method based on the ratio of disease names, respectively. In addition, the system proposed in this paper, UI/UX was configured conveniently via the feedback of actual users. Each system screen has a title and a description of the screen at the top, and was configured to display a user to conveniently view the information on the disease. The information of the disease searched based on the calculation method proposed above displays images and disease names of similar diseases. The system’s environment is implemented for use with a web browser based on a pc environment and a web browser based on a mobile device environment. 상의 특징인 색상, 모양, 질감 등을 이용해 상을 검색하는 연구들은 많이 진행되어 왔다. 한 작물의 질병 상 과 련된 연구들도 진행되고 있다. 농업 장에서 재배되는 작물에 발생한 질병을 확인하는데 도움이 되기 해 본 논문에서는 시설원 작물의 질병 상을 이용한 유사도 기반 작물 질병 검색 시스템을 제안한다. 제안하는 시스템 은 단일 Descriptor를 사용하지 않고, 조합 Descriptor를 통해 기존 비 상의 유사도 검색 성능을 높고 유사 도 검색 결과를 가독성 높게 사용자에게 제공하기 해 가치 기반 산출방법을 용했다. 본 논문에서는 총 13개 의 개별 Descriptor를 이용해 조합을 진행했다. 조합 Descriptor를 이용해 6개 작물의 질병에 해 유사도 검색을 진행했고 작물별로 평균 accuracy가 높은 조합 Descriptor를 선정해 유사도 검색에 사용했다. 검색된 결과는 병명 의 비율을 기반으로 한 산출방법과 가치를 기반으로 한 산출방법을 사용해 백분율로 나타냈다. 병명의 비율을 기 반으로 한 산출방법은 질의 상과 유사도 검색에 사용되는 상의 수가 많은 병명이 1순로 출력되는 문제이 있다. 이를 해결하기 해 가치를 기반으로 한 산출방법을 사용했다. 작물의 병명별 테스트 상을 두 가지 산출 방법에 용해 검색 성능을 측정했다. 작물의 질병별로 두 가지 산출방법에 해 검색 성능 값의 평균을 비교한 결 과 고추, 사과 작물에서는 병명의 비율을 기반으로 한 산출방법의 성능이 가치를 기반으로 한 산출방법의 성능보 다 평균 약 11.89%의 높은 성능 결과를 보다. 국화, 딸기, 배, 포도 작물에서는 가치를 기반으로 한 산출방법 이 병명의 비율을 기반으로 한 산출방법의 성능보다 평균 약 20.34%의 높은 성능 결과를 보다. 한 본 논문에 서 제안하는 시스템의 UI/UX는 실제 사용자의 피드백을 통해 편리하게 구성했다. 시스템의 화면마다 상단에 제목 과 설명을 출력했고 사용자가 질병의 정보를 보기 편리하게 화면을 구성했다. 검색된 질병의 정보는 에서 제안한 산출방법을 토로 유사한 질병의 상과 병명을 출력한다. 시스템의 환경은 PC 환경 기반의 웹 라우와 모바일 디바이스 환경 기반의 웹 라우를 통해 사용할 수 있도록 구했다.

농작물 질병분류를 위한 전이학습에 사용되는 기초 합성곱신경망 모델간 성능 비교

윤협상(Hyoup-Sang Yoon),정석봉(Seok-Bong Jeong) 한국산업경영시스템학회 2021 한국산업경영시스템학회지 Vol.44 No.3

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.

윤작이 감자 수량, 토양 화학성 및 미생물 활성에 미치는 영향

김유경,강호준,양상호,오한준,이신찬,강성근,김형신,Kim, Yu-Kyoung,Kang, Ho-Jun,Yang, Sang-Ho,Oh, Han-Jun,Lee, Shin-Chan,Kang, Seong-Keun,Kim, Hyoung-Sin 한국유기농업학회 2012 韓國有機農業學會誌 Vol.20 No.4

The objective of this study was to determine crop rotation effects on potato yield, soil chemical and microbiological properties from a short-term field experiment from 2010 to 2011 in Jeju Island, Korea. Potato cropping systems included continuous and rotation sequences of soybean (Glycine max(L.) Merr.), barley (Hordeum vulgare var. hexastichon), rapeseed (Brassica napus L.) and broccoli (Brassica oleracea var. italica). Crop rotations increased the yields of potato from 31% to 52% compared with continuous potato. Marketable yield of potato was highest under soybean plus rapeseed rotation by $20.97MT\;ha^{-1}$ and lowest under continuous cropping by $11.95MT\;ha^{-1}$. The incidence and severity of scab disease was significantly lower in tubers from crop rotation with soybean plus barley. Differences in marketable tuber yields among rotations were associated with potato scab disease. Especially, incidence and severity of potato scab were strongly correlated with soil pH, exchangeable calcium, and bacteria population of the soil. Crop rotations significantly increased soil pH, available phosphate, exchangeable K and Ca, especially in crop rotations with soybean plus barley or rapeseed. Soil microbial biomass C of crop rotations with soybean plus barley or rapeseed, was also significantly higher compared with monoculture. In conclusion, crop rotation may decrease the incidence of soil-born pathogen by increasing soil chemical properties and soil microbial biomass. Overall, potato crop productivity was generally maintained in rotations that contained soybean plus barley or rapeseed but declined under continuous cropping system. 본 시험은 윤작처리가 감자의 상품수량과 토양 화학성 및 미생물 활성에 미치는 영향을 평가하고 알맞은 작부체계를 확립하기 위하여 시험을 수행하였다. 작부체계내 콩과 보리를 도입하여 2년 2기작으로 감자를 윤작하였을 경우 총수량과 상품수량은 대조구인 연작구 대비 각각 31% 및 53% 정도 증가하였으며, 콩과 유채를 도입하여 2년 2기작으로 감자를 윤작하였을 경우 총수량과 상품수량은 각각 57% 및 75% 정도 증가하였다. 그리고 감자의 상품수량에 미치는 요인간 상관관계를 분석한 결과 상품수량은 토양미생물 활성, 특히 Biomass C 및 세균 밀도가 높을수록 증가하였고, 토양 pH, 유효인산, 치환성 칼슘함량이 높을수록 세균 밀도가 증가하였다. 따라서 윤작을 통해서 토양 화학성을 개량하고 미생물 밀도 및 활성을 증가시키면, 토양병 발생을 억제하고 감자의 생산성을 높일 수 있을 것으로 판단된다.

벼 도열병 발생 탐지 및 확산 모니터링을 위한 기상자료, 위성영상, 드론영상의 공동 활용

류재현,안호용,이경도 한국농림기상학회 2023 한국농림기상학회지 Vol.25 No.4

The representative crop in the Republic of Korea, rice, is cultivated o ver extensive areas every year, which resulting in reduced resistance to pests and diseases. One of the major rice diseases, rice blast disease, can lead to a significant decrease in yields when it occurs on a large scale, necessitating early detection and effective control of rice blast disease. Drone-based crop monitoring techniques are valuable for detecting abnormal growth, but frequent image capture for potential rice blast disease occurrences can consume significant labor and resources. The purpose of this study is to early detect rice blast disease using remote sensing data, such as drone and satellite images, along with weather data. Satellite images was helpful in identifying rice cultivation fields. Effective detection of paddy fields was achieved by utilizing vegetation and water indices. Subsequently, air temperature, relative humidity, and number of rainy days were used to calculate the risk of rice blast disease occurrence. An increase in the risk of disease occurrence implies a higher likelihood of disease development, and drone measurements perform at this time. Spectral reflectance changes in the red and near-infrared wavelength regions were observed at the locations where rice blast disease occurred. Clusters with low vegetation index values were observed at locations where rice blast disease occurred, and the time series data for drone images allowed for tracking the spread of the disease from these points. Finally, drone images captured before harvesting was used to generate spatial information on the incidence of rice blast disease in each field.

인삼 논재배 연작지에서 윤작물 재배가 토양화학성, 토양 미생물상 및 2년생 인삼의 생육에 미치는 영향

이성우,박경훈,이승호,장인복,Mei Lan Jin 한국약용작물학회 2016 한국약용작물학회지 Vol.24 No.4

Background: Crop rotation plays an important role in improving soil chemical properties, minimizing the presence of disease pathogens, and assists in neutralizing autotoxic effects associated with allelochemicals. Methods and Results: Five rotation crops of sudan grass, soybean, peanut, sweet potato, and perilla were cultivated for one year with an aim to reduce yield losses caused by repeated cropping of ginseng. In 2-year-old ginseng grown in the same soil as a previous ginseng crop, stem length and leaf area were reduced by 30%, and root weight per plant was reduced by 56%. Crop rotation resulted in a significant decrease in electrical conductivity, NO3, and P2O5 content of the soil, whereas organic matter, Ca, Mg, Fe, Cu, and Zn content remained-unchanged. Soil K content was increased following crop rotation with sudan grass and peanut only. Rotation with all alternate crops increased subsequent ginseng aerial plant biomass, whereas root weight per plant significantly increased following crop rotation with perilla only. A significant positive correlation was observed between root rot ration and soil K content, and a significant negative correlation was observed between ginseng root yield and the abundance of actinomycetes. Crop rotation affected the soil microbial community by increasing gram negative microbes, the ratio of aerobic microbes, and total microbial biomass whereas decreases were observed in actinomycetes and the ration of saturated fatty acids. Conclusions: In soil exhibiting crop failure following replanting, crop rotation for one year promoted both soil microbial activity and subsequent ginseng aerial plant biomass, but did not ameliorate the occurrence of root rot disease.

이모작 적응 조생 복합내병충성 고 바이오매스 사료용 벼 ‘조우’

안억근(Eok-Keun Ahn),원용재(Yong-Jae Won),이점호(Jeom-Ho Lee),정응기(Eung-Gi Jeong),박향미(Hyang-Mi Park),홍하철(Ha-Cheol Hong),장재기(Jae-Ki Chang),현웅조(Ung-Jo Hyun),정국현(Kuk-Hyun Jung),이정희(Jeong-Heui Lee) 한국육종학회 2021 한국육종학회지 Vol.53 No.2

‘Jowoo’, an early maturing, high biomass yielding rice (Oryza sativa L.) cultivar with high feed value and multiple disease-insectresistance, was developed for winter forage crop-whole crop silage (WCS) rice double cropping. It was derived from a cross between a highbiomass and multiple disease resistant line ‘Suweon519’ and an early flowering, high yielding Tong-il type variety ‘Gaya’ that harbored Bph3and BPH26, conferring brown planthopper (Bph) broad resistance. This cultivar had a growth period of approximately 106 days from seedingto heading, 97 cm culm length, 24 cm panicle length, 12 panicles per hill, 120 spikelets per panicle, and 1,000-grain weight of 22.1 g asbrown rice in the central plain region of Suwon. This early flowering rice variety was tolerant to premature germination and cold stresses,especially showed higher grain fertility than ‘Nokyang’ at a cold-water (17℃) irrigated screening nursery. Further, ‘Jowoo’ exhibited high lodgingresistance in the yellowing ripening stage, an optimal period for harvest. In addition, ‘Jowoo’ was resistant to leaf and neck blast, bacterialblight (race K1, K2, K3, K3a), rice stripe virus, and small Bph and was also moderately resistant to Bph. Its average dry matter yield forthree years reached 18.2 MT/ha, the highest among early flowering rice varieties and 19% higher than that of ‘Nokyang’. This cultivar had72.2% total digestible nutrients, which is also higher than that of ‘Nokyang’. In the Korean peninsula, ‘Jowoo’ grows well in the centraland southern plains and can be harvested 30 days after heading. This helps improve its feeding value and digestion rate in livestock (GrantNo. 8133).

Surveys on Disease Occurrence in Major Horticultural Crops in Kangwon Alpine Areas

함영일,권민,김점순,서효원,안재훈,Hahm, Young-Il,Kwon, Min,Kim, Jeom-Soon,Seo, Hyo-Won,Ahn, Jae-Hoon The Korean Society of Plant Pathology 1998 Plant Pathology Journal Vol.14 No.6

강원 고랭지의 대표적 지대인 평창의 대관령, 홍천의 내면 그리고 횡성의 둔내지역에서 재배되고 있는 주요 원예작물에 대해 1994년$\sim$1997년까지 4년에 걸쳐 실시한 병해조사의 결과는 다음과 같다. 1) 배추와 무에서는 순무모자이크바이러스(TuMV)가 '92년부터 고랭지에 다발생하여 '94년과 '96년에 피해가 심하였으며, 이 바이러스병과 함께 무름병의 발생이 심한 편이었으며, 최근에는 무사마귀병의 피해고 '96년부터 점증 추세에 있으며 수년 전까지 피해가 심했던 뿌리 마름병은 최근 발생이 지극히 저조한 점이 특이하다. 거의 모든 채소작물에서 무름병, 잿빛곰팡이병, 노균병, 흰가루병, 잘록병, 검은무늬병 등의 발생이 심한 것이 특징인데 저지대와 달리 고랭지 기후환경과 무관하지 않았을 것이다. 한편 조사기간 중 새로운 미기록종으로는 셀러리의 바이러스(BBWV)와 잿빛곰팡이병(Botrytis cinerea), 메론의 점무늬병(Cercospora citullina), 딸기의 흰가루병(Erysiphe polygori), 양상추의 흰가루병(Erysiphe cichoraceaium) 등이 밝혀졌으며 기주 미기록으로는 파슬리의 무사마귀병(Plasmodiophora brassicae)이 밝혀진 것이 특이하다. 2)여러 가지 화훼작물에는 주로 바이러스병, 잿빛곰팡이병, 시들음병, 흰무늬병 등의 발생이 많았으며 특히 자생 나리류에 흰무늬병(Cercospora spp.) 발생이 심하였다. 미기록종으로는 꽃도라지의 시들음병(Fusarium oxysporum f. sp. eustomae), 바이러스병(BBWV, CMV), 균핵병(Sclerotinia sclerotiorum), 잿빛곰팡이병(B. cinerea), 카네이션의 반점병(Cladosporium echinulatum), 용담의 시들음병(Fusarium oxysporum), 잎마름병(Alternavia dianthi) 등과 Phytoplasma에 의한 스타티스의 빗자루병, 리아트리스의 빗자루병, 자생나리류의 흰무늬병(Cercospora sp.), 스토크의 TuMV 등이 동정되었다. These surveys were conducted to check the occurrence of disease in various horticultural crops in alpine areas, especially Daekwallyong areas, Pyongchang-Gun, and Hyeongseong-Gun in Kangwon province. TuMV on Chinese cabbage was one of the most serious diseases, especially in 1994 and 1997. The incidence of soft rot and clubroot has been increased gradually. Brittle root rot on Chinese cabbage was significantly decreased. Soft rot, gray mold, downy mildew, powdery mildew, bottom rot and Alternaria leaf spot were the common diseases on most vegetable crops. Gray mold (Botrytis cinerea) on celery, cercospora leaf spot (Cercospora sp.) on melon, powdery mildew (Erysiphe cichoracearum) on lettuce, and clubroot (Plasmodiophora brassicae) on parsley are newly found in Korea. The most common and predominant diseases were viruses, especially CMV, TMV, TuMV, BBMV, and gray mold, wilts, and cercospora leaf spot on many flowers in alpine areas. Fusarium oxysporum f. sp. eustomae causing wilting on lisianthus (Eustoma grandiflorum), Turnip mosaic virus causing mosaic and color breaking on stock, Cercospora spp. causing cercospora leaf spot on various wild lily, Cladosporum echinulatum causing leaf spot on carnation, and phytoplasma causing witches' broom on statice (Limonium sinuatum) and blazing star (Liatris spp.) were newly found during these surveys in Korea.

Feeding the extra billions: strategies to improve crops and enhance future food security

Stamm, Petra,Ramamoorthy, Rengasamy,Kumar, Prakash P. The Korean Society of Plant Biotechnology 2011 Plant biotechnology reports Vol.5 No.2

The ability to feed an expanding world population poses one of the greatest challenges to mankind in the future. Accompanying the increased demand for food by the expected nine billion inhabitants of Earth in 2050 will be a continual decrease in arable land area, together with a decline in crop yield due to a variety of stresses. For these formidable challenges to be met, future crops should not only by high-yielding, but also stress-tolerant and disease-resistant. In this review, we highlight the importance of genetic engineering as an indispensable tool to generate just such future crops. We briefly discuss strategies and available tools for biotechnological crop improvement and identify selected examples of candidate genes that may be manipulated so that current biological maxima in yield may be surpassed by comfortable margins. Future prospects and the necessity for basic research aimed at identifying novel target genes are also discussed.

유통과정 작물의 질병 분류를 위한 YOLO 기반 객체탐지

김가영,곽명섭,이상민 한국로지스틱스학회 2022 로지스틱스연구 Vol.30 No.6

Most agricultural products in South Korea are cultivated in the open field and delivered to customers through the distribution channels. It is important to ensure that crops are not susceptible to pests and diseases during delivery. In this paper, we propose to detect diseases of crops in earlier time by using YOLOv4, which is a real-time object detection algorithm, to ensure the freshness of crops. We compared the proposed detector with the YOLOv3-based detector in terms of the accurate performance. The experimental results addressed that the proposed method clearly detects the types and stages of chili pepper diseases. We expect that the proposed approach can be used for a fast and accurate freshness monitoring system.