What happened
The Rural Development Administration runs the National Crop Pest Management System (NCPMS). Gyeonggi-do Agricultural Research and Extension Services collected images for its deep-learning diagnosis of leek and cucumber diseases from 2021 to 2024. Staff photographed symptoms on farms in Anseong, Pyeongtaek and Hwaseong with smartphones, confirmed the pathogens by culture and DNA methods and labelled the images in the crop pest management program. The dataset held 12,236 images for three leek diseases and six cucumber diseases, including standard images by disease stage. To check the dataset's accuracy, the institute ran field demonstrations in May and October 2024 in which farmers photographed disease symptoms with their own smartphones. The smart pest and disease diagnosis app, built with a deep-learning algorithm, analysed the photos. The institute's 2024 research report, published in June 2025, states that the diagnosis rate was 90% or more in both rounds. It adds that the app can diagnose quickly and accurately on farms. The institute supplied the training data. The report gives no counts of farms or photos behind the rate, and no outcome for the farms is reported.
Sources
-
Report, by an evaluator that the makers or users commissioned.
실증을 통해 학습된 딥러닝 데이터세트의 정확성과 재현성을 검토하였고, 1차, 2차 모두 90% 이상의 진단율을 확인하였다
The trials demonstrated that the developed app exhibited high accuracy and reproducibility, enabling fast and accurate disease diagnosis on farms, and is expected to make a significant contribution to pest and disease management.
Details
AI system
- Workflow
-
Staff photographed disease symptoms on leek and cucumber farms from 2021 to 2024, confirmed the pathogens and labelled 12,236 images for deep learning. Farmers photographed symptoms with their own smartphones in two demonstration rounds, in May and October 2024. The app analysed the photos with a deep-learning algorithm.
Follow-up
- Conditions for replication
-
The disease must be one of the three leek or six cucumber diseases in the training data. Photos must show symptoms as they appear on farms, since the dataset was built by disease stage and plant part.
Entry details
- Sector
-
Agriculture / crop disease diagnosis.
- Purpose
-
The Rural Development Administration's crop pest management program and the provincial institute built the deep-learning dataset so that the smart pest and disease diagnosis app can diagnose leek and cucumber diseases from farmers' photos without an expert.
31 more questions not yet checked
model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, AI contribution, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.
Cite this benefit
Peter H. Vartanian. (2026). 스마트 병해충 진단 앱 tops 90% on leek and cucumber farm photos (AIBD-1199). Artificial Intelligence Benefits Database, edition of 2 November 2026. https://petervartanian.xyz/aibd/records/AIBD-1199.html (archived at https://doi.org/10.7910/DVN/AADSII, version 9.0)