Method for propagating image label based on minimal cost path

Abstract

The invention relates to a method for propagating an image label based on a minimal cost path; according to the method, the defect that an undirected graph becomes into a sparse directed graph as the graph only has a plurality of adjacent nodes for each node is resolved by constructing the graph and reconstructing the sparse matrix of the graph; data is quickly classified by the improved minimum spanning tree algorithm so as to reduce the time complexity of the algorithm and achieve the purpose of completely sorting all data. According to the method, the optimal paths of image data samples for obtaining the labels are searched from unmarked image data samples, thus ensuring that each image data sample can be effectively classified, and also ensuring that only one label is propagated for each point; construction of the sparse neighbor matrix is improved, and the problem that propagation of the image data sample labels is incomplete and incorrect due to the neighbor matrix is reduced; as propagation of the algorithm is intercepted among the different labels, the algorithm can be applied to large-scale multi-classification data with multiple labels.

Description

Technical field [0001] The invention belongs to the cross-research fields of pattern recognition, artificial intelligence and image processing, and relates to a method for image label propagation based on a minimum cost path. Background technique [0002] With the development of information technology, information networks are flooded with more and more data information, such as massive data texts and high-resolution images. How to correctly organize and quickly utilize massive data has become a research hotspot in the field of machine learning. Manual labeling of samples in massive amounts of data is costly, so labeled data is often too scarce and precious. For example, for some more complex images, it is more difficult for ordinary users to interpret the content, and experts are usually required to complete it. For example, a large number of image data samples may be stored in a medical database, and doctors mark the possible causes of the image according to the image data sa...

AI Technical Summary

Problems solved by technology

[0003] However, the high time complexity seriously restricts the application of graph-based semi-supervised classification methods in the field of large-scale data classification.

Although the Minmax Label Propagation (MMLP) algorithm proposed by Kim and Choi in 2014 reduces the time complexity of the algorithm to O(n), but because the MMLP algorithm does not solve the problem of the sparse similarity matrix The connectivity problem in the middle picture makes it impossible to completely classify the disconnected sparse similarity matrix composed of large-scale data.

Images