Design and Implementation of AlexNet Flower Classification System from the Perspective of Deep Learning

Abstract

In the field of computer vision, image classification is a fundamental and important task. Traditional image classification methods rely on manually designed feature extractors such as SIFT and HOG, but these methods have limited effect when processing complex image data. With the rise of deep learning technology, especially the rapid development of convolutional neural network (CNN), the accuracy and efficiency of image classification have been significantly improved. AlexNet As a classic model of deep convolutional neural network, it has become a research hotspot in the field of computer vision since it achieved excellent results in the ImageNet large-scale visual recognition competition in 2012. Flower classification is mainly based on botanical and morphological characteristics, which helps scientists and horticulturists to better understand and study different species of flowers. The classification of flowers also helps people to better understand the growth habits, ecological environment and use of flowers, which is of great significance for horticultural cultivation and biodiversity conservation. The experiment mainly uses Alexnet network through the flower image classification, flower five types, through the training set on the flower image feature extraction and model training, and validation in the test set to complete the experiment, the experiment of Alexnet network layer of the corresponding changes, in order to reduce the purpose of calculating complexity, at the same time keep the training accuracy and previous network.