Insect order classification using faster R-CNN Deep learning algorithm

Insects are the most crowded animal group in terms of class of the arthropod branch and in terms of species and taxa. In fact, 750 thousand identified in insects, the estimated number of species is around 1.5 million. In particular, destructions cause them to disappear without specifying forest fires. Therefore, biodiversity and academic studies are extremely important. Also, knowing the benefits and harms of insects at the team level and having more information will contribute to the studies especially in agriculture. On the other hand, thanks to the recognition of insects, by imitating nature in technological fields, it is very useful and inspiration for human beings who develop robot technologies, sensors and smart systems in various fields. In recent years, thanks to the methods and methods developed in artificial intelligence, deep learning, image processing techniques, detection, classification and identification of objects have been carried out. In many areas, researchers are using these libraries to develop algorithms for the detection, classification and identification of objects of different properties. Scientifically; Scientists dealing with entomology determine teams based on these criteria. Having too many criteria can lead to time loss and misdiagnosis. In addition, there is no software to check the diagnoses made. In this study, a model for the detection and classification of insects on Insects was proposed with the Faster R-CNN Deep Learning method using the Tensorflow library. The software created is presented to the literature with a system that enables fast detection of insects at the team level.