MENU

Notices

WAGRI-API for AI image diagnosis of diseases and insect pests released by the National Agricultural Research Institute of Japan - New service with additional target crops and small-scale illustrated book of diseases and insect pests

  • Mar 28, 2022
Notices

The target crops of WAGRI1) diagnostic service for pests and diseases have been expanded from 4 crops (tomato, cucumber, strawberry, eggplant) to 8 crops (peach, grape, sweet pepper, soybean, potato, pumpkin, chrysanthemum, onion ) and now 12 crops. A new service that combines the AI diagnosis service and WAGRI's small illustrated book of diseases and pests, which provides basic information on diseases and pests, will be made available for a fee on April 28 as " WAGRI-API3) for AI diagnosis of diseases and pests by the National Institute of Agrobiological Sciences. This service will enable rapid diagnosis of diseases and insect pests in agricultural fields.

For more information on this press release, please click here (jump to the official website of the National Agricultural Research Organization).

1) WAGRI
WAGRI is a platform for data linkage developed under the Strategic Innovation Program (SIP) "Next Generation Agriculture, Forestry and Fisheries Creation Technology" of the Cabinet Office (https://wagri.net, connection confirmed in March 2022), In addition to various paid data from private companies, various public data (land and map information, soil, weather, market conditions, etc.) and services (crop growth and yield prediction, etc.) related to agriculture are available for commercial use.

2) AI
Artificial intelligence (AI) is a computer system that does not simply make computers perform calculations, but also performs understanding, inference, and suggestions that until now could only be performed by humans. In the field of AI for image identification, neural networks, which mimic the information transfer of human neurons, have been used for many years, but as computer performance has improved, deep learners with many layers of pseudo-neurons (Convolutional Neural Network However, as computer performance improved, the development of deep learners (Convolutional Neural Networks) with many layers of pseudo-neurons dramatically improved recognition capabilities.

3) API (Application Programming Interface)
API is a communication method to connect different applications. By connecting to the "WAGRI-API" and using the discrimination function, private companies can provide unique disease and pest diagnosis services to end-user farmers.
First, the "WAGRI-API" inputs images sent to WAGRI from farmers via a service program of a private company to the "Image Discriminator" in the National Institute of Agrobiological Sciences. The "image discriminator" then outputs the results of disease/insect identification corresponding to the input images to WAGRI, which returns the output results to the service program of the private sector. Finally, the results of the disease and insect identification are provided to the farmer by the private operator. In this way, the NRI AI disease and insect damage image diagnosis WAGRI-API is provided to private operators such as ICT vendors as an API on WAGRI. Therefore, this system is not suitable for direct use by the general public.