Automatic image semantic annotation method based on scale learning and correlated label dissemination

Abstract

The invention relates to an automatic image semantic annotation method based on scale learning and correlated label dissemination, which comprises the following steps: firstly, the global and partial feature descriptor of each image is extracted after the image library is read; the feature descriptor is sent to a model based on a structured support vector machine for learning the distance scale between the images, actually the Mahalanobis distance; a model about the internal relation between key words is built; the learned Mahalanobis distance is embedded in a built label dissemination model so as to obtain the confidence degree score of each key word belonging to the image to be labeled; and a threshold value is set for the confidence degree score of each key word, and the key words of which the scores are higher than the threshold value are distributed to the images to be labeled, thereby completing labeling. The learning algorithm model based on the structured support vector machine can effectively solve the measuring problem of similarity between the images, the internal relation between the key words is fully excavated through the embedded-type correlated label dissemination model, and the accuracy of the image annotation and image retrieval is effectively improved.

Description

technical field [0001] The invention relates to the technical field of image retrieval and automatic image labeling, in particular to an automatic image semantic labeling method based on scale learning and associated label propagation. Background technique [0002] In recent years, with the rapid development of Internet technology, multimedia images, and storage devices, the number of images that users come into contact with has grown explosively. How to quickly and effectively search for the information users need from massive data is an important research topic, which is also the research content in the field of image retrieval. [0003] The initial image retrieval was based on text-based image retrieval (TBIR for short), which first described the content of each image manually with a series of keywords, and then classified the images according to these keywords. Indexing is established, so that the image retrieval problem is transformed into the matching problem of text ...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to address the deficiencies of the existing methods, and propose an automatic image semantic labeling method based on scale learning and associated label propagation, which overcomes the problem of only using predefined distances such as Euclidean distance to measure the distance between images in the prior art. The defect of semantic similarity, while fully mining the correlation between keywords, makes the result of annotation more accurate and effective

Images