DEVELOPMENT OF SOFTWARE FOR COLLECTION OF PHENOTYPIC DATA OF SUNFLOWER SEEDS

Abstract

The main direction in sunflower breeding is the improvement of seed quality and industrial characteristics, which are important for the consumer. Phenotypic evaluation can be used to assess seed quality. Phenotyping allows for a formalized approach to assessment that includes a list of traits, a form of precise description, and means to measure those characteristics. The purpose of the presented research is the development of software for the study of morphological features of sunflower seeds based on photo images, namely metric characteristics, shape and color of seeds. Methodology. The developed programs are based on accurate mathematical calculation, which ensures the accuracy of the created descriptions. Image analysis software products are developed using the Python programming language. The calculation program was developed in the PyCharm Community Editor 2022.2.4 environment using the following libraries: os, cv2, imutils, numpy, time, Pool. Cv2-for image processing. The study was conducted on a collection of sunflower seed photographs, which contains 3,000 photographs of 250 sunflower lines and was created by specialists of the Institute of Oil Crops of the National Academy of Agrarian Sciences of Ukraine. Developments can be used for the analysis of photo images of seeds, which is made on a stationary device with deterministic parameters in laboratory conditions. The scientific novelty is that, for the first time, programs were developed for the analysis of the image of sunflower seeds, taking into account the biological diversity of the seeds. For the first time, they made it possible to describe the qualitative characteristics of seeds in the form of tables of quantitative indicators. Conclusions. As a result of the research, a model was developed for measuring the metric characteristics and shape of the seed, as well as software, which consists of programs for recognizing the metric characteristics of the seed, shape and striped color of the seed. The obtained results are an important component for selection work to create new high-yielding hybrids, varieties and lines of sunflower. Now the development is used by breeders and geneticists of the Institute of Oil Crops of the National Academy of Agrarian Sciences of Ukraine in the course of scientific selection work.