Butterfly detection and classification techniques: A review
Rumana Yasmin, Aparna Das, Liton Jude Rozario, Md. Ezharul Islam
Abstract
Automatic identification of insects from images is gaining interest and popularity day by day. Several methods and systems were developed to identify insects accurately. Among all the species of insects, the task of different butterfly species identification and classification is very complex, laborious, and time-consuming. Butterfly identification and classification is not only important in the sector of computer vision but also equally important in the field of entomology to preserve the butterfly species. Various methods essentially image processing, computer vision, traditional machine learning, neural networks, deep learning, transfer learning, etc. have been employed until now to detect and classify butterflies. The purpose of this review paper is to portray a summary of various proposed methods of butterfly species identification as well as classification strategies so that it might be helpful for future studies to develop new methods.