Small fishing boat classification and recognition method based on GASF temporal graph coding and EMPViT model

Jiaqi Deng ¹, Xin Liu ², Gang Du ²

⁰Southwest Institute of Electronic Technology, Chengdu, 610036, China. djqjecky@163.com.

Scientific Reports +

|

February 7, 2025

View abstract on PubMed

Summary

This study introduces a novel method for identifying small fishing vessels using Gramian Summation Angular Field (GASF) image encoding and an enhanced Efficiency MPViT (EMPViT) model. The approach achieves high accuracy in complex environments, improving ship management.

Area Of Science

Marine Engineering
Computer Vision
Artificial Intelligence

Background

Effective management of small fishing vessels is crucial.
Current identification methods face challenges in complex environments.
Accurate classification of small fishing vessels is needed for enhanced maritime surveillance and resource management.

Purpose Of The Study

To propose a robust classification and recognition method for small fishing vessels.
To address the complexities of the recognition environment for small fishing boats.
To improve the accuracy and performance of small fishing vessel identification systems.

Main Methods

Utilized a high-precision laser sensor for one-dimensional contour data acquisition.
Employed polynomial fitting for contour shape delineation.
Encoded contour data into two-dimensional time series images using Gramian Summation Angular Field (GASF).
Applied an enhanced Efficiency MPViT (EMPViT) model for classification and identification.

Main Results

The GASF encoding and EMPViT model achieved a peak accuracy of 99.98% for small fishing vessel classification.
Ablation experiments confirmed the superiority of the EMPViT model over traditional CNN and ViT models.
Demonstrated high accuracy and performance in challenging recognition scenarios.

Conclusions

The proposed GASF sequence diagram coding and EMPViT model offer a highly effective solution for small fishing vessel identification.
This method significantly enhances the accuracy and performance compared to existing approaches.
The findings contribute to improved ship management and maritime surveillance capabilities.

Keywords:

Classification and recognition of small fishing boats EMPViT deep learning model GASF time series images One-dimensional time series coding Polynomial fitting

Related Concept Videos

Modeling and Similitude

Scaled modeling is a fundamental technique in engineering, enabling the study of large and complex systems by creating smaller, manageable replicas that recreate critical characteristics of the original. In hydrology and civil infrastructure, for example, scaled models of dams help analyze water flow, turbulence, and pressure. This method allows for accurate predictions of real-world behavior within a controlled environment, significantly reducing the cost and time involved in full-scale...