Physics Maths Engineering
Rong Lan,
Haowen Mi,
Na Qu,
Feng Zhao,
Haiyan Yu,
Lu Zhang
Abstract Although evidence c-means clustering (ECM) based on evidence theory overcomes the limitations of fuzzy theory to some extent and improves the capability of fuzzy c-means clustering (FCM) to express and process the uncertainty of information, the ECM does not consider the spatial information of pixels, which makes it to be unable to effectively deal with noise pixels. Applying ECM directly to image segmentation cannot obtain satisfactory results. This paper proposes a robust evidence c-means clustering combining spatial information for image segmentation algorithm. Firstly, an adaptive noise distance is constructed by using the local information of pixels to improve the ability to detect noise points. Secondly, the pixel’s original, local and non-local information are introduced into the objective function through adaptive weights to enhance the robustness to noise. Then, the entropy of pixel membership degree is used to design an adaptive parameter to solve the problem of distance parameter selection in credal c-means clustering (CCM). Finally, the Dempster’s rule of combination was improved by introducing spatial neighborhood information, which is used to assign the pixels belonging to the meta-cluster and the noise cluster into the singleton cluster. Experiments on synthetic images, real images and remote sensing SAR images demonstrate that the proposed algorithm not only suppress noise effectively, but also retain the details of the image. Both the segmentation visual effect and evaluation indexes indicate its effectiveness in image segmentation.
ECM is a clustering method based on evidence theory that improves upon fuzzy c-means clustering (FCM) by better handling uncertainty in data. However, traditional ECM doesn’t consider spatial information, making it less effective for noisy images.
Traditional ECM doesn’t account for the spatial relationships between pixels, which makes it sensitive to noise and less effective for image segmentation tasks.
The proposed method enhances ECM by incorporating spatial information, adaptive noise detection, and improved Dempster’s rule of combination. This makes it more robust to noise and better at preserving image details.
The method uses:
Dempster’s rule is improved by incorporating spatial neighborhood information. This helps assign pixels from meta-clusters and noise clusters to singleton clusters, improving segmentation accuracy.
Adaptive weights balance the influence of a pixel’s original, local, and non-local information in the clustering process. This enhances the method’s robustness to noise while preserving image details.
The method uses the entropy of pixel membership degrees to design an adaptive parameter. This solves the problem of selecting distance parameters in credal c-means clustering (CCM), making the method more flexible and accurate.
The method was tested on synthetic images, real images, and remote sensing SAR images. Results showed that it effectively suppresses noise while retaining image details, outperforming traditional ECM and other clustering methods.
The method:
This method is useful for:
While deep learning methods require large labeled datasets, this method is more interpretable and doesn’t rely on extensive training. It is particularly effective for noisy images where deep learning might struggle without sufficient data.
Future research could explore:
Show by month | Manuscript | Video Summary |
---|---|---|
2025 April | 4 | 4 |
2025 March | 67 | 67 |
2025 February | 53 | 53 |
2025 January | 56 | 56 |
2024 December | 53 | 53 |
2024 November | 56 | 56 |
2024 October | 64 | 64 |
2024 September | 86 | 86 |
2024 August | 74 | 74 |
2024 July | 50 | 50 |
2024 June | 30 | 30 |
2024 May | 29 | 29 |
2024 April | 50 | 50 |
2024 March | 15 | 15 |
Total | 687 | 687 |
Show by month | Manuscript | Video Summary |
---|---|---|
2025 April | 4 | 4 |
2025 March | 67 | 67 |
2025 February | 53 | 53 |
2025 January | 56 | 56 |
2024 December | 53 | 53 |
2024 November | 56 | 56 |
2024 October | 64 | 64 |
2024 September | 86 | 86 |
2024 August | 74 | 74 |
2024 July | 50 | 50 |
2024 June | 30 | 30 |
2024 May | 29 | 29 |
2024 April | 50 | 50 |
2024 March | 15 | 15 |
Total | 687 | 687 |