MPCM-Net: A multiscale network that integrates partial attention convolution with mamba for ground-based cloud image segmentation
Author Identifier (ORCID)
Jianxin Li: https://orcid.org/0000-0002-9059-330X
Abstract
Ground-based cloud image segmentation is a critical research domain for photovoltaic (PV) power forecasting. Current deep learning (DL) approaches primarily focus on encoder-decoder architectural refinements. However, existing methodologies exhibit several limitations: 1) they rely on dilated convolutions for multiscale context extraction, yet fail to leverage interchannel interoperability and partial feature efficacy; 2) implementations of attention-based feature enhancement frequently compromise the equilibrium between accuracy and throughput; and 3) the decoder modifications often fail to re-establish global interdependencies among hierarchical local features, thereby constraining inference efficiency. To mitigate these challenges, we propose MPCM-Net, a multiscale network that integrates partial attention convolutions with Mamba architectures to enhance segmentation accuracy. Specifically, the encoder incorporates a multiscale partial attention convolution (MPAC), which comprises: 1) a multiscale partial convolution (MPC) block with partial channel module (ParCM) and partial spatial module (ParSM) that facilitating global spatial interaction across multiscale cloud formations and 2) a multiscale partial attention (MPA) block combining partial attention module (ParAM) and ParSM to extract discriminative features with reduced computational complexity. On the decoder side, a multiscale Mamba block (M2B) is employed to mitigate contextual loss through a spatial-semantic hybrid domain (SSHD) that maintains linear complexity while enabling deep feature aggregation across spatial and scale dimensions. Furthermore, we introduce and release a dataset incorporating complex-scale variations, radiative properties, and color attributes (CSRC), which is a clear-label, fine-grained segmentation benchmark designed to overcome the critical limitations of existing public datasets. Extensive empirical analysis on CSRC demonstrates the superior performance of MPCM-Net over state-of-the-art (SOTA) methods, achieving an optimal balance between segmentation accuracy and inference speed. The dataset and source code will be available at https://github.com/she1110/CSRC.
Keywords
Fine-grained segmentation dataset, ground-based cloud image, multiscale network, partial attention convolution
Document Type
Journal Article
Date of Publication
1-1-2026
Volume
64
Publication Title
IEEE Transactions on Geoscience and Remote Sensing
Publisher
IEEE
School
School of Business and Law
Funders
National Natural Science Foundation of China (62206085) / Innovation Capacity Improvement Plan Project of Hebei Province (22567603H) / Interdisciplinary Postgraduate Training Program of Hebei University of Technology (HEBUT-Y-XKJC-2022101)
Copyright
subscription content
Comments
Niu, P., She, J., Cai, T., Zhang, Y., Zhang, P., Gu, J., & Li, J. (2026). MPCM-Net: A multiscale network that integrates partial attention convolution with mamba for ground-based cloud image segmentation. IEEE Transactions on Geoscience and Remote Sensing, 64, 1–16. https://doi.org/10.1109/TGRS.2026.3666092