Wake detection model construction method based on frequency domain multi-scale perception

By constructing a multi-scale frequency domain sensing wake detection model, the problem of low AUV wake detection accuracy under complex sea surface backgrounds was solved, achieving efficient wake feature extraction and improved detection accuracy.

CN121936320AActive Publication Date: 2026-04-28CHINA UNIV OF PETROLEUM (EAST CHINA)
5 Cites 1 Cited by

Patent Information

Application Number
CN202610407736.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-31
Publication Date
2026-04-28
Estimated Expiration
2046-03-31

AI Technical Summary

Technical Problem

Existing technologies struggle to effectively detect AUV wakes against complex sea surface backgrounds, resulting in low detection accuracy and susceptibility to environmental interference, particularly in the insufficient capture of weak wake texture information.

Method used

A wake detection model based on frequency domain multi-scale perception is constructed. Simulated SAR images are generated by fusing the sea surface background and internal wave wakes into a composite model. The wake detection model is optimized by combining the RT-DETR model with a multi-scale texture information selection network, a frequency domain multi-scale attention perception module, and a golden cudgel convolution downsampling module.

Benefits of technology

It significantly improves the ability to extract high-frequency textures and fine-grained structures against complex sea surface backgrounds, increases detection accuracy and reduces computational overhead, achieving the best balance between detection accuracy and computational efficiency.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a wake detection model construction method based on frequency domain multi-scale perception, which relates to the technical field of ocean observation, is used for capturing weak wake features to carry out sea surface internal wave wake detection, and comprises the steps of constructing a wake image data set, constructing a wake detection model by taking an RT-DETR model as a baseline model, and constructing a multi-scale texture information selection network. A backbone network of the baseline model is replaced, a frequency domain multi-scale attention perception module is constructed, an AIFI module of the baseline model is replaced, a golden cudgel convolution down-sampling module is adopted to replace a standard convolution down-sampling module of the baseline model, training and optimization are performed on the wake detection model based on the wake image data set, and a target detection result is obtained. According to the method, the wake detection model AUVRT-DETR is constructed, the detection precision is improved while the advantage of light weight is kept, the optimal balance is achieved between the detection precision and the calculation efficiency, and the model parameter quantity and the calculation overhead are remarkably reduced while the highest detection precision is achieved.
Need to check novelty before this filing date? Find Prior Art