The paper introduces the Channel-Aware U-Shaped Mamba (CU-Mamba) model, which integrates a dual State Space Model (SSM) framework into the U-Net architecture for image restoration. CU-Mamba addresses the limitations of CNNs and Transformer-based models, such as inadequate long-range dependency modeling and high computational costs, by employing a Spatial SSM module for global context encoding and a Channel SSM component to preserve channel correlation features. Both modules operate with linear computational complexity relative to the feature map size. Extensive experiments on various datasets, including image denoising and deblurring, demonstrate that CU-Mamba outperforms existing state-of-the-art methods, highlighting the importance of integrating both spatial and channel contexts in image restoration. The model's effectiveness is further validated through ablation studies, which show that both spatial and channel SSM modules contribute significantly to the model's performance.The paper introduces the Channel-Aware U-Shaped Mamba (CU-Mamba) model, which integrates a dual State Space Model (SSM) framework into the U-Net architecture for image restoration. CU-Mamba addresses the limitations of CNNs and Transformer-based models, such as inadequate long-range dependency modeling and high computational costs, by employing a Spatial SSM module for global context encoding and a Channel SSM component to preserve channel correlation features. Both modules operate with linear computational complexity relative to the feature map size. Extensive experiments on various datasets, including image denoising and deblurring, demonstrate that CU-Mamba outperforms existing state-of-the-art methods, highlighting the importance of integrating both spatial and channel contexts in image restoration. The model's effectiveness is further validated through ablation studies, which show that both spatial and channel SSM modules contribute significantly to the model's performance.