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.

광교, 동북태, 강림, 육우3호, 은대두등과 같은 대리장려품종이 바이러스에 의하여 심하게 이병되었다. 이 병은 주로 모자익병의 발생이 많은 강원, 경기지방에서 발병이 심하였으나 모자익병이 심하지 않은 전남 등 남부지방에서도 발병되고 있다. 병징으로 보아 tobacco ringspot virus에 의한 대두의 피해와 유사한 것으로 보였으나 지표식물검정과 혈청검정에 의하여 조사한 결과 모두 부정적이었으며 대두품종에 따른 이병정도의 상이, 품종과 접종원에 의한 병징의 변이가 많았다. 이병주에서 분리되는 병징형은 Mottling과 necrosis였으며 지금까지의 연구결과 이 대두병해는 모자익바이러스(SMV)의 계통 내지는 tobacco ringspot virus 이외의 두류바이러스의 복각감염에 의한 것으로 생각할 수 있으나 SMV의 계통에 의한 피해일 가능성이 더욱 유력시되고 있다.

This study is designed to assess the priority order of the chemicals to cause to generate occupational diseases in order to understand the fundamental data required for the preparation of health protective measure for the workers dealing with chemicals. The 41 types of 51 ones of chemicals to cause to generate the national occupational diseases were selected as the study objects by understanding their domestic use or not, and their occupational diseases' occurrence or not among 110,608 types of domestic and overseas chemicals. To assess their priority order the sum of scores was acquired by understanding the actually classified condition based on a perfect score of physical riskiness(90points) and health toxicity(92points) as a classification standard by GHS, the priority order on GHS riskiness assessment, GHS toxicity assessment, GHS toxic․riskiness assessment(sum of riskiness plus toxicity) was assessed by multiplying each result by each weight of occupational disease's occurrence. The high ranking 5 items of chemicals for GHS riskiness assessment were turned out to be urethane, copper, chlorine, manganese, and thiomersal by order. Besides as a result of GHS toxicity assessment the top fives were assessed to be aluminum, iron oxide, manganese, copper, and cadium(Metal) by order. On the other hand, GHS toxicity․riskiness assessment showed that the top fives were assessed to be copper, urethane, iron oxide, chlorine and phenanthrene by order. As there is no material or many uncertain details for physical riskiness or health toxicity by GHS classification though such materials caused to generate the national occupational diseases, it is very urgent to prepare its countermeasure based on the forementioned in order to protect the workers handling or being exposed to chemicals from health.