Image Classification Algorithm Based on Improved AlexNet in Cloud Computing Environment

In the cloud computing environment, the traditional classification algorithms often ignore the feature relationship between images, which leads to unstable classification process, poor accuracy of classification results and other problems, which can not achieve the ideal classification effect. An image classification algorithm based on improved AlexNet is proposed and designed. After preprocessing the collected images, such as normalization, mean value and standardization, the convolutional nerve is introduced to train the features of the standard images. On this basis, the image classification algorithm model based on improved AlexNet is established. Through the optimization training of the classification model, the high-level semantic features of the image are extracted, and the process of image classification and calculation is realized. The experiments show that the improved AlexNet image classification algorithm improves the accuracy and stability of image classification, and has good effectiveness and robustness, which provides a good reference for the development of network technology.