Side-scan sonar sunken ship target unsupervised detection method based on selective search algorithm

A side-scan sonar and search algorithm technology, applied in computing, computer parts, image data processing, etc., can solve problems such as difficult to guarantee accuracy, insufficient expert image database, and low computing efficiency, so as to improve detection accuracy and simplify Measuring the effects of process and efficiency

Active Publication Date: 2020-01-10
JIANGSU OCEAN UNIV
View PDF3 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

As a kind of underwater acoustic equipment, side-scan sonar is used in complex and changing marine environments, which determines that there are not enough expert image libraries to choose from, and the target features extracted from images obtained under different sea conditions are used for target detection in unfamiliar sea conditions Time accuracy is difficult to guarantee
[0005] The other is a method based on unsupervised learning, which usually requires a certain mathematical model. When these models are applied to side-scan sonar waterfall images with a large amount of data, they often have the defect of low computational efficiency.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Side-scan sonar sunken ship target unsupervised detection method based on selective search algorithm
  • Side-scan sonar sunken ship target unsupervised detection method based on selective search algorithm
  • Side-scan sonar sunken ship target unsupervised detection method based on selective search algorithm

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0037] Embodiment 1, a side-scan sonar sunken ship target non-supervised detection method based on selective search algorithm, its steps are as follows:

[0038] Step 1: Combine the last peak method, the principle of seabed line symmetry and the principle of gradual change of seabed topography to implement accurate seabed tracking;

[0039] Step 2: Filter and denoise the water column area and the seabed area respectively; considering that the sunken ship is the sinking target and the water column area is an invalid area, simply mark the image pixel values ​​of the port and starboard water column areas as 0 to form an overall area; The area is based on Gaussian filtering to achieve filtering and denoising;

[0040] Step 3: Implement side-scan sonar waterfall image segmentation based on selective search strategy; the specific process is as follows:

[0041] 1) Pre-segment the side-scan sonar waterfall image based on simple k-means clustering to form a pre-segmented area; the pa...

Embodiment 2

[0058] Embodiment 2: A side-scan sonar sunken ship target non-supervised detection method based on selective search algorithm, the steps are as follows:

[0059] 1: Selective search strategy:

[0060] 1) Pre-segment the pre-processed striped side-scan sonar waterfall image based on k-means classification to form an initial segmented area R={r 1 ,r 2 ,...,r n};

[0061] 2) Calculate the similarity s(r) between two adjacent areas in texture and shape i ,r j ), to obtain the similarity set S={s(r i ,r j ),...};

[0062] 3) The region r corresponding to the value with the largest similarity in the set S i and r j merged into r t , and remove s(r i ,r j ); continue to calculate r t Similarity with neighboring regions, and r t Join the segmented region set R;

[0063] 4) Repeat step 3) until At this time, the subset in the obtained set R is the final segmentation area; due to the introduction of the wreck shape and the multi-fractal spectrum wide texture feature tha...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a side-scan sonar sunken ship target unsupervised detection method based on a selective search algorithm. The method comprises the following steps: preprocessing a side-scan sonar stripe waterfall image; dividing the strip waterfall image into a water column area, a target area (shadow area) and a pure seabed background area on the basis of prior knowledge according to thebasic characteristics of the side-scan sonar, and segmenting the side-scan sonar waterfall image into the areas on the basis of a selective search strategy; defining a plurality of similarity measures, calculating the similarity measure of each region, and taking a weighted value as a final measure value; and outputting a shipwreck target detection result. According to the method, the unsuperviseddetection of the shipwreck target in the large-data-volume side-scan sonar strip waterfall image is effectively realized, a sample image is not needed, a shipwreck identification model does not needto be constructed, and the detection process and efficiency of the underwater shipwreck target are greatly simplified.

Description

technical field [0001] The invention relates to the technical field of side-scan sonar, in particular to a non-supervised detection method for side-scan sonar sunken ship targets based on a selective search algorithm. Background technique [0002] Side Scan Sonar (SSS) images have important application value in the detection and identification of mine-like objects, submarine cold seeps, underwater shipwrecks and other underwater targets. Target detection can be achieved based on the time-domain Ping section data, but the Ping echo intensity is significantly affected by complex ocean noise, and it is difficult to effectively and accurately apply the target Ping section model based on the laboratory. [0003] In current technology, there are two main target detection methods based on airspace image data: [0004] One is a method based on supervised learning, which extracts the shape, texture, grayscale, shape and other features of the reference target image in the known image...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/11G06T7/48G06T5/00G06K9/62
CPCG06T7/11G06T7/48G06T5/002G06T2207/10132G06T2207/20024G06T2207/20221G06F18/23213G06F18/22
Inventor 王晓韩友美杨敬华张博宇李书东鲍天宇
Owner JIANGSU OCEAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap