Novel automatic scorpion-detection and -recognition system based on machine-learning techniques

Abstract All species of scorpions can inject venom, some them even with the possibility killing a human. Therefore, early detection and identification are essential to minimize scorpion stings. In this paper, we propose novel automatic system for recognition using computer vision machine learning (ML) approaches. Two complementary image-processing techniques were used proposed method accurately reliably detect presence scorpions. The first is based on fluorescent characteristics when exposed ultraviolet light, second shape features Also, three models ML algorithms image classification compared. particular, found in La Plata city (Argentina): Bothriurus bonariensis (of no sanitary importance), Tityus trivittatus , confluence (both importance) have been researched local binary-pattern histogram algorithm deep neural networks transfer (DNNs TL) data augmentation TL DA) A confusion matrix receiver operating characteristic curve evaluate quality these models. results obtained show that model DNN DA most efficient at simultaneously differentiating between (for health security) T. biological research purposes).