Texture feature and shape feature fusion-based underwater sound image feature extraction method

Abstract

The invention discloses a texture feature and shape feature fusion-based underwater sound image feature extraction method. The method mainly comprises the following key steps of: carrying out image segmentation; carrying out boundary extraction; generating an element matrix; calculating a grayscale-element co-occurrence matrix; and obtaining five feature quantities. According to the method, graphical features of underwater sound images of undersea sediment are set about, and the feature extraction of underwater sound images of undersea sediments is realized by applying and combining the boundary extraction and the grayscale-element co-occurrence matrix and utilizing boundary shape features of the underwater sound images of the undersea sediments and grayscale relevance. The method is capable of ensuring translation, rotation and zooming invariance, is insensitive to noise, and is capable of describing closed areas and well completing feature extraction for non-closed areas, so as to realize indirect recognition and extraction on the basis of graphical features of underwater sound images of undersea sediments.

Description

technical field [0001] The invention belongs to the technical field of image processing, and specifically relates to an underwater acoustic image feature extraction method based on the fusion of texture features and shape features. Background technique [0002] The principle of image feature extraction is image recognition, which is actually an auxiliary process of classification. In order to identify the category of an image, it is necessary to distinguish it from other images of different categories, which requires that the selected features not only be able to describe the image well, but more importantly, be able to distinguish different categories well. Image. Select those image features with small differences between images of the same type and large differences between images of different categories, which are called the most discriminative features. [0003] Generally speaking, there are many ways to describe image features. For example, according to the size of th...

AI Technical Summary

Problems solved by technology

[0008] The purpose of the present invention is to solve the problems in the prior art that are affected by the resolution, illumination, reflection and shape complexity, resulting in errors in classification or a large amount of calculation. Acoustic image feature extraction method, this method starts from the study of graphic features of hydroacoustic images of different seabed substrates, and completes the recognition and extraction of hydroacoustic image features of seabed substrates such as sand, clay, and gravel

Afterwards, the gray-element co-occurrence matrix of the boundary image with the original gray value is calculated, which solves the shortcomings of using a certain method alone, such as large amount of calculation, complex expression, poor retrieval effect, etc., and can well combine statistical methods and Structural analysis methods are organically combined

Images