Desenvolvimento de aplicação web para identificação de plantas alimentícias não convencionais a partir de imagens

Abstract

This project aimed to develop an application capable of identifying unconventional edible plants (Panc) through image capture and providing users with basic information and ways of consuming the identified plant. Software development tools, the Python programming language, and images of the species captured in the UTFPR - Pato Branco campus area were used for the application’s development. The generated model is capable of identifying up to 40 Panc species and can be supplemented with more information to expand its identification capabilities. The use of this technology can yield positive results in teaching botany and increasing the general population’s knowledge of Panc. Additionally, it can serve as a aid in reducing food and nutritional insecurity.