Scene recognition method based on layered Gaussian hybrid model

Abstract

The invention discloses a scene recognition method based on a layered Gaussian hybrid model. The method comprises the following steps: extracting the layered Gaussian hybrid model corresponding to all images in a training image library and obtaining lie algebra descriptors of the layered Gaussian hybrid model; receiving images to be sorted and extracting the lie algebra descriptors of the layered Gaussian hybrid model of the images to be sorted; obtaining the distances from the lie algebra descriptors of the layered Gaussian hybrid model of the images to be sorted to the lie algebra descriptors of the layered Gaussian hybrid model of all images in the training image library; and sorting the images to be sorted according to the obtained distances by a minimum mid-value distance sorter and returning sorting results. According to the method provided by the invention, character representation of the layered Gaussian hybrid model of the images is optimized, so that the scene recognition performance is greatly improved.

Description

technical field [0001] The invention belongs to the field of scene recognition, and more specifically relates to a scene recognition method based on a layered Gaussian mixture model. Background technique [0002] The scene recognition method mainly includes extracting image feature descriptors and selecting an appropriate classifier. Extracting image feature descriptors is the core technology of scene recognition. The availability of feature descriptors for images is a key issue. Now image feature descriptors mainly include statistical histogram, area covariance and single Gaussian type. The statistical histogram is the approximation of the probability density function. The histogram has the following disadvantages: it is very sensitive to the number of noise and feature quantization; the dimension and channel number of the histogram show exponential growth; because the histogram distribution is no longer For vector space, if the measurement method of vector space is used...

AI Technical Summary

Problems solved by technology

[0006] Aiming at the defects of the prior art, the purpose of the present invention is to provide a scene recognition method based on a layered Gaussian mixture model, which aims to solve the problem that the structure information of the original space in the existing scene recognition method is easy to be destroyed, and there is no factor The problem of a large amount of information redundancy caused by data correlation between different dimensions of the vector

Images