Power transformation equipment image super-resolution reconstruction method and system based on lightweight network

The lightweight network-based super-resolution image reconstruction method addresses the challenge of capturing detailed features in power equipment images by using improved NSTB and channel attention mechanisms, achieving efficient and high-quality image reconstruction suitable for real-time monitoring and maintenance.

CN119671857APending Publication Date: 2025-03-21STATE GRID ANHUI ELECTRIC POWER CO LTD ELECTRIC POWER SCI RES INST +2
1 Cites 0 Cited by

Patent Information

Application Number
CN202411753756.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-02
Publication Date
2025-03-21

AI Technical Summary

Technical Problem

Existing methods for super-resolution image reconstruction of power equipment images face challenges in capturing detailed features due to reliance on large network models, limited hardware resources, and the complexity of power equipment backgrounds, which results in low-resolution images that are difficult to analyze, especially in real-time monitoring and maintenance scenarios.

Method used

A lightweight network-based super-resolution image reconstruction method using a trained model that incorporates improved Noise Suppression Transformer Blocks (NSTB), channel attention mechanisms, and Patch Merging layers to effectively capture local and global features, reducing computational complexity and enhancing image clarity and detail.

Benefits of technology

The method effectively increases the resolution and detail of power equipment images, providing clearer visual information for monitoring and maintenance, while being efficient enough to operate in resource-constrained environments, improving image quality and reducing computational load.

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

Abstract

The invention discloses a power transformation equipment image super-resolution reconstruction method and system based on a lightweight network, and the method comprises the steps: obtaining a high-resolution power transformation equipment image through employing a trained image super-resolution reconstruction model; wherein the modeling and training process of the image super-resolution reconstruction model comprises the following steps of: collecting a high-resolution image of power transformation equipment, and generating a low-resolution image by using the high-resolution image; establishing an image super-resolution reconstruction model; and sending the generated low-resolution image as input data into an image super-resolution reconstruction model for training, learning a mapping relation between the low-resolution image and the corresponding high-resolution image, and gradually adjusting parameters of the low-resolution image so as to effectively generate the high-resolution image. According to the invention, clearer visual information can be provided for monitoring of a power system.
Need to check novelty before this filing date? Find Prior Art