A detection method based on feature information fusion of multi-view target bright spots

Abstract

The present invention provides a detection method based on multi-view target bright point feature information fusion. The method comprises: 1, using N sensors distributed on a target space to obtain target echo signals and perform pre-processing; 2, using a CLEAN feature extraction method to extract target feature information that meets conditions; 3, calculating each-order central moment as statistical features of bright points; 4, taking the obtained the statistical features of the target bright points obtained as input of a vector machine classifier, performing estimation of incident anglesof detection signals to obtain estimation values of N detection signal incident angles; 5, estimating echo signals of each sensor in the same incident angle; 6, performing fusion of the estimated echo signals after registration; and 7, taking changing conditions of each bright point intensity of signals after fusion as judgment basis to perform judgment of the detection target. The detection method based on multi-view target bright point feature information fusion can be applied to a multi-base sonar system located at different target viewing angles and can be applied to a single-base sonar system of a relative target movement.

Description

technical field [0001] The invention relates to an underwater acoustic signal processing method, in particular to a detection method based on the fusion of multi-viewpoint target bright spot feature information. Background technique [0002] In recent years, in the use of information fusion for target detection, the main work has focused on the application of target features and vector machine classifiers. For example, in the literature (Lee S J, Choi I ​​S, Cho B, et al. Performance enhancement of target recognition using feature vector fusion of monostatic and bistatic radar [J]. Progress In Electromagnetics Research, 2014, 144: 291-302.), by extracting different The scattering characteristics of the bright spots of the target under the viewing angle are then classified and recognized by the neural network, and the recognition performance is better than that of a single platform. However, since this method directly combines the target features extracted by sensors under d...

AI Technical Summary

Problems solved by technology

However, first of all, according to the target strength of common sonar targets given in the book "Practical Sonar Engineering" (A.D.Waite, Wang Deshi. Practical Sonar Engineering [M]. Electronic Industry Press, 2004.), the different types of targets The range of scattering intensity is similar; secondly, the target scattering intensity is closely related to the propagation loss in the active sonar detection process, and the propagation loss is affected by the complex marine environment; thirdly, such as submarine reefs, fish schools, etc. can produce echoes, and the target intensity may also be close to the target strength

In summary, all of the above factors may lead to false alarms in the target recognition method based on the target strength.

Images