Desenvolvimento de algoritmo de k-nearest neighbors (vizinhos mais próximos) para classificação de imagens de organismos do plâncton.

This study aimed to develop a k-nearest neighbors (KNN)-based algorithm for the automated recognition of planktonic organisms from digital images. Image datasets were used to train and evaluate the classifier, and its performance was assessed based on classification accuracy. The KNN approach achieved accuracy rates comparable to those obtained through conventional image identification by non-specialist observers, suggesting its potential as an alternative computational tool for plankton ecological analyses.