# PM-MCD: A network combining pyramid feature extraction and multi-scale attention fusion for multiclass change detection

## TL;DR

The paper introduces PM-MCD, a new model for detecting changes in satellite images that outperforms existing methods by using pyramid features and multi-scale attention.

## Contribution

The novel PM-MCD model combines pyramid feature extraction and a multi-scale attention fusion module for improved multiclass change detection.

## Key findings

- PM-MCD achieves 99.4% overall accuracy on the WHU-CD dataset.
- The model outperforms CNN- and Transformer-based methods in terms of mIoU and F1 scores.
- The MSSC module enhances detection of small-scale change regions.

## Abstract

Multiclass change detection in remote sensing images plays a vital role in remote sensing applications. However, the existing methods still have the problem of subtle changes missed. In this paper, we propose a model named PM-MCD, which consists of a VMamba-based pyramid feature extraction encoder for remote sensing images and a multi-scale information aggregation decoder based on MLP and MSSC module, enabling efficient multiclass change detection in remote sensing images. In addition, we propose a multi-scale attention fusion module, MSSC, to enhance the model’s ability to recognize small-scale change regions. Experimental results show that, on the WHU-CD, Landsat-SCD, and CNAM-CD datasets, our model outperforms existing CNN- and Transformer-based methods, achieving 99.4/96.77/90.86% overall accuracy (OA), 90.18/82.27/68.50% mean intersection over union (mIoU), and 91.44/89.88/79.86% F1 scores.

•An end-to-end multiclass change detection model for remote sensing images•Multi-scale and attention mechanism are used to enhance the model’s performance•Achieve superior performance on multiple datasets

An end-to-end multiclass change detection model for remote sensing images

Multi-scale and attention mechanism are used to enhance the model’s performance

Achieve superior performance on multiple datasets

Remote sensing; Computer modeling