Method for searching image based on amplitude phase hybrid modeling

Abstract

The invention discloses a method for searching an image based on amplitude phase hybrid modeling. The method comprises the following steps: firstly, carrying out PDTDFB conversion on a texture image, respectively carrying out amplitude calculation on an imaginary part and a real part of a high-frequency sub-band obtained after conversion; secondly, modeling a statistic model for the amplitude by applying a Weibull distribution function to obtain a shape parameter gamma and a dimension parameter phi; modeling a statistic model for a relative phase by applying Vonn distribution function to obtain a position parameter mu and a dimension parameter lambda; and finally searching an image by applying gamma, phi, mu and lambda which serve as image features. Experiment results indicate that the method adopts parameters gamma, phi, mu and lambda as features of each image to effectively reduce the dimension of features and reduce the time distributed for similarity calculation, and has a relatively high average searching rate and relatively low time complexity.

Description

technical field [0001] The invention relates to a content-based image retrieval method, in particular to an amplitude-based image retrieval method that can effectively reduce the dimension of features, reduce the time allocated for similarity calculation, and have a higher average retrieval accuracy and lower time complexity. Image retrieval methods for value-phase mixture modeling. Background technique [0002] Today, with the explosion of information technology, Internet culture has penetrated into people's daily life. The application of picture information and multimedia technology has prompted people to urgently need excellent algorithms and technologies to filter the required information. Therefore, how to retrieve and classify large amounts of information more efficiently and accurately The source of digital images is a hot issue of common concern, and content-based image retrieval (CBIR) is one of the effective technologies to solve the problem. [0003] Compared wit...

AI Technical Summary

Problems solved by technology

At present, although there are many image retrieval methods based on color and contour texture information, there are still many difficulties in the classification of contour texture features, resulting in low average retrieval rate and high time complexity, and it is difficult to achieve all contours with a single method. Texture Features for Classification

Images