Abstract

Insects are everywhere and at any time and are capable of causing many problems, including in homes, farms, and public health. Since the traditional method of identifying and classifying insect species can be time-consuming and prone to errors, there is an urgent need for more efficient and accurate methods. Machine learning has emerged as an insect detection and classification tool, using algorithms to process massive amounts of data and extract relevant features. This comprehensive survey aims to provide an overview of the latest insect detection and classification developments using machine learning in modern applications. The survey covers various topics, including common types of insects, challenges in detecting and classifying insect species, and the techniques and algorithms used in the field. Therefore, automating this process may reduce expenses and improve accuracy and scalable analytics. The survey also provides a comprehensive overview of different machine-learning algorithms for insect detection and classification, including supervised and unsupervised learning algorithms such as support vector machines, k-nearest neighbors, random forests, convolutional neural networks, clustering, and anomaly detection. The survey highlights the advantages and limitations of each algorithm and its respective applications. Therefore, this research proposes a questionnaire considering the following computer science digital search databases: IEEE, Science Direct, Scopus, Springer Link, and Web of Science. Current trends and obstacles in discovery are discussed in all papers published between 2018 and 2022. The results show how deep learning strategies such as convolutional neural networks, enhanced feature extraction, and ignoring hash can be used in practice.