Image classification method and device, equipment and storage medium

Abstract

The invention relates to an image classification method and device, equipment and a storage medium, and the method comprises the steps: preprocessing a to-be-processed original image to acquire a corresponding input image; performing feature extraction on the input image by adopting a pre-training model to obtain a plurality of feature data; wherein each piece of feature data corresponds to different image layers of the input image; performing dimension reduction on the extracted feature data, so that a feature set after dimension reduction can be obtained; connecting the feature data of different dimensions in the feature set in series to form a feature change curve; and classifying an original image by utilizing the trained classification network model according to the characteristic change curve. A pre-training model is directly adopted for extraction when feature extraction is carried out, a large number of samples do not need to be manufactured for training the network model for feature extraction, dependence on the training samples is effectively reduced, the image classification process is simplified, and overall efficiency of image classification is improved.

Description

technical field [0001] The present application relates to the technical field of image processing, and in particular to an image classification method and device, device and storage medium. Background technique [0002] With the development of deep learning technology, deep learning has become the mainstream of image processing. In related technologies, when processing images, a supervised learning method is usually used. When using supervised learning methods for image processing, one is: use CNN convolutional neural network to extract image features for classification, input the original image into the network, label data as supervision, and the learning process is convolutional neural network shallow network to extract image low-level features , such as: edges, lines, textures, etc., and pass them layer by layer to obtain deep image features, update the network weights through backpropagation of labeled results, continuously learn the best classification network, and fin...

AI Technical Summary

Problems solved by technology

Although the image classification can be realized after the neural network is trained by the supervised learning method, the above-mentioned method has a strong dependence on the sample data, and a large amount of training sample data needs to be collected, which makes the image classification process more cumbersome and time-consuming.

Images