Image clustering method and system

Abstract

The invention discloses an image clustering method and system, which comprises the following steps: obtaining an image data set according to an image to be processed; processing the image data set to obtain a target projection matrix, and all projection directions of the target projection matrix The sum is accumulated to an extreme value; performing dimensionality reduction processing on the image data set according to the target projection matrix to obtain an image subspace; performing clustering processing on the image subspace to obtain clustering information of the image to be processed. The image clustering method of the present invention realizes dimension reduction processing before image clustering by finding the target projection matrix whose sum of all projection directions is an extreme value, which not only retains the structure of data features, but also ensures that each projection direction finds the most Optimal solution, so as to obtain the global optimal projection direction, avoid the feature redundancy problem of high-dimensional image data, and improve clustering accuracy and clustering efficiency.

Description

technical field [0001] The invention relates to the field of machine learning, in particular to an image clustering method and system. Background technique [0002] In recent years, with the development and popularization of the Internet, the amount of data such as images, videos, and texts and the feature vectors representing the data are increasing, that is to say, the dimensionality of the data is increasing. In order to utilize these massive data, it is necessary to cluster these high-dimensional data quickly and effectively. [0003] However, these high-dimensional big data contain a lot of redundant information, and the direct clustering effect is not good. Therefore, dimensionality reduction is used before clustering to reduce the data dimension and improve the effective utilization of data features. The commonly used principal component analysis (PCA) method does not fully preserve the structure of the data features; the trace ratio method (Trace Ratio) can make fu...

AI Technical Summary

Problems solved by technology

The commonly used principal component analysis (PCA) method does not fully preserve the structure of the data features; the trace ratio method (Trace Ratio) can make full use of the label information of the training set, but in actual use, it may get poor projection directions

Images