A Study on Classification and Detection of Small Moths Using CNN Model

Currently, there are many limitations to classify images of small objects. In addition, such as error detection due external factors, and is also a disadvantage that it difficult accurately distinguish between various This paper uses convolutional neural network (CNN) algorithm recognize object very moths obtain precise data images. A convolution used for image classification, the classified transformed into learn topological structure image. To improve accuracy classification reduce loss rate, parameter finding fast-optimal point set by pixel preprocessor. As result this study, we applied capturing moths. Experimental results showed was more than 90%.