Automatic classification method of side scan sonar image targets based on transfer learning and depth learning
A side-scan sonar and deep learning technology, applied in instruments, character and pattern recognition, computer parts, etc., can solve problems such as too little sample data, inability to apply deep learning solutions, etc., to increase the number of basic features, reduce The effect of optimizing the number of parameters and improving the classification accuracy
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment approach
[0037] A method for automatically classifying objects in side-scan sonar images based on transfer learning and deep learning. The specific implementation method mainly includes the following steps:
[0038] 1. Obtain a suitable conventional optical image dataset and its segmentation annotations. Example images and segmentation annotations in the dataset are as follows: figure 2 and image 3 shown.
[0039] 2. Extract the contour of the corresponding image in the data set by using the segmentation annotation, and then perform preprocessing on the image inside and outside the contour, change the gray scale range, add noise, etc., and the preprocessed data set is called the source domain data set. Preprocessing results such as Figure 4 shown.
[0040] 3. Select a convolutional neural network structure for classification, and use the source domain data set for sufficient training, save the trained network, and name it as the source domain classification network.
[0041] 4. ...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com