App for smartphone for detecting fungus diseases of plant leaves

Abstract

The symptoms and biophysical processes occurring in garden strawberry plants when they are affected by the dominant type of disease (up to 80%) caused by pathogenic fungi have been described. The ineffectiveness of the visual assessment of the degree of damage to strawberry diseases by a conventional 5-point scale or as a percentage of the leaf plate area affected by fungi, with the involvement of qualified specialists, has been shown. To create diagnostic tools that allow early detection of fungal diseases of garden strawberries, one of the methods of computer vision was proposed by counting image pixels in the space of color channels of red, green and blue (R, G, B), which makes it possible to determine the degree of fungal diseases affecting an individual plant leaf. The algorithm includes capturing an image with a digital camera by focusing on a plant leaf placed on a substrate with a uniform background providing a contrasting selection of the object; converting a color image to black and white; dividing the image between areas with necrotic spots and healthy areas of the plant leaf by masking and removing pixels; counting the number of pixels in these two areas and calculating their ratio. Information about a computer software for determining the degree of damage to a strawberry leaf by garden fungal diseases has been given. Java programming language (operating system Android Studio 3.4.1) was used as a language for the development of the logical part of the information system. In order to build a graphical interface, the software facilitating the development and integration of various modules of the LibGDX software project was used. The proposed algorithm is implemented for a personal computer and can be installed on a smartphone in the form of a software application, with the help of which any agricultural producer can carry out early diagnosis of fungal plant diseases.

Plant phenotyping relevance

イチゴ葉の病斑面積比を画像処理で定量化し、植物病害の程度を推定する手法とスマートフォンソフトウェアの開発が中心であるため含める。

Code and data availability

The article describes a smartphone/PC app for detecting fungal disease on strawberry leaves using RGB pixel counting, but provides no public dataset, image collection, code repository, or download link for the authors' software. The only URL besides the DOI is the generic libGDX developer guide, a third-party framework

No evidence-backed public reproduction asset is currently recorded.