en

Project:

DEVELOPMENT OF A MOBILE APPLICATION BASED ON CONVOLUTIONAL NEURAL NETWORKS FOR DISEASE IDENTIFICATION BY PLANT IMAGE

Intro

It is said that a face can say a lot about human’s health. It is even believed that specific areas of the face are linked to specific internal issues. This is where a very natural question appears - If it works for people, why can’t it work for plants?

We believe that people are not an exception. Trust us, pictures of plants can be a lot more than just beautiful.

How does it work?

An innovation in the course of the project is a mobile application based on neural networks for the identification of diseases according to the image of the plant.

The user-sent image of the plant, photographed by a smart device, is processed by an information system that operatively and accurately identifies plant (winter and spring wheat, barley, potatoes, beans and peas) diseases and their prevalence in plants using the developed object recognition classifier. The main task of innovation is to detect plant diseases as early and accurately as possible and to select appropriate protection measures, thus ensuring crop protection and the sustainable use of chemicals.

Why is this needed?

The aspect of plant protection measures is one of the most important areas in the European agricultural sector. The EU emphasizes that plant protection must be based on the control of diseases and pathogens, giving priority to those that pose the least risk to human health and the environment. Also, protecting against diseases and pathogens must reconcile economic benefits with safety for humans and nature.

What is the benefit?

Users of the app can use resources and plant protection products more rationally, protect crops in a timely manner and thus increase farm productivity and profitability. What is more, optimal use of chemical plant protection products reduces risks to human health and the negative impact of farming on the environment.


+370 630 78989
Draugystes str. 20, LT-89167 Mazeikiai, Lithuania

Įmonės kodas: 300635091
PVM mokėtojo kodas: LT100004103016
Atsiskaitomoji sąskaita: LT837290000008467114