{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/superpca-a-superpixelwise-pca-approach-for","title":"SuperPCA: A Superpixelwise PCA Approach for Unsupervised Feature Extraction of Hyperspectral Imagery","arxiv_id":"1806.09807","date":"2018-06-26","proceeding":null,"authors":["Junjun Jiang","Jiayi Ma","Chen Chen","Zhongyuan Wang","Zhihua Cai","Lizhe Wang"],"abstract":"As an unsupervised dimensionality reduction method, principal component\nanalysis (PCA) has been widely considered as an efficient and effective\npreprocessing step for hyperspectral image (HSI) processing and analysis tasks.\nIt takes each band as a whole and globally extracts the most representative\nbands. However, different homogeneous regions correspond to different objects,\nwhose spectral features are diverse. It is obviously inappropriate to carry out\ndimensionality reduction through a unified projection for an entire HSI. In\nthis paper, a simple but very effective superpixelwise PCA approach, called\nSuperPCA, is proposed to learn the intrinsic low-dimensional features of HSIs.\nIn contrast to classical PCA models, SuperPCA has four main properties. (1)\nUnlike the traditional PCA method based on a whole image, SuperPCA takes into\naccount the diversity in different homogeneous regions, that is, different\nregions should have different projections. (2) Most of the conventional feature\nextraction models cannot directly use the spatial information of HSIs, while\nSuperPCA is able to incorporate the spatial context information into the\nunsupervised dimensionality reduction by superpixel segmentation. (3) Since the\nregions obtained by superpixel segmentation have homogeneity, SuperPCA can\nextract potential low-dimensional features even under noise. (4) Although\nSuperPCA is an unsupervised method, it can achieve competitive performance when\ncompared with supervised approaches. The resulting features are discriminative,\ncompact, and noise resistant, leading to improved HSI classification\nperformance. Experiments on three public datasets demonstrate that the SuperPCA\nmodel significantly outperforms the conventional PCA based dimensionality\nreduction baselines for HSI classification. The Matlab source code is available\nat https://github.com/junjun-jiang/SuperPCA","url_abs":"http://arxiv.org/abs/1806.09807v2","url_pdf":"http://arxiv.org/pdf/1806.09807v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"superpca-a-superpixelwise-pca-approach-for","repo_url":"https://github.com/junjun-jiang/SuperPCA","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"dimensionality-reduction","task_name":"Dimensionality Reduction"},{"task_slug":"classification","task_name":"General Classification"}],"methods":[{"method_slug":"pca","method_name":"PCA"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}