{"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/pansharpening-via-detail-injection-based","title":"Pansharpening via Detail Injection Based Convolutional Neural Networks","arxiv_id":"1806.08898","date":"2018-06-23","proceeding":null,"authors":[],"abstract":"Pansharpening aims to fuse a multispectral (MS) image with an associated\npanchromatic (PAN) image, producing a composite image with the spectral\nresolution of the former and the spatial resolution of the latter. Traditional\npansharpening methods can be ascribed to a unified detail injection context,\nwhich views the injected MS details as the integration of PAN details and\nband-wise injection gains. In this work, we design a detail injection based CNN\n(DiCNN) framework for pansharpening, with the MS details being directly\nformulated in end-to-end manners, where the first detail injection based CNN\n(DiCNN1) mines MS details through the PAN image and the MS image, and the\nsecond one (DiCNN2) utilizes only the PAN image. The main advantage of the\nproposed DiCNNs is that they provide explicit physical interpretations and can\nachieve fast convergence while achieving high pansharpening quality.\nFurthermore, the effectiveness of the proposed approaches is also analyzed from\na relatively theoretical point of view. Our methods are evaluated via\nexperiments on real-world MS image datasets, achieving excellent performance\nwhen compared to other state-of-the-art methods.","url_abs":"http://arxiv.org/abs/1806.08898v1","url_pdf":"http://arxiv.org/pdf/1806.08898v1.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":"pansharpening-via-detail-injection-based","repo_url":"https://github.com/XiaoXiao-Woo/PanCollection/tree/dev/UDL/pansharpening/models/DiCNN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"pansharpening","task_name":"Pansharpening"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/pansharpening-on-full-worldview-3","task":"Pansharpening","dataset":"Full WorldView-3 PanCollection","model":"LAGConv","rank_in_archive_order":1,"of":1,"metrics":{"D_lambda":"0.0368","D_s":"0.0418","HQNR":"0.9230"},"uses_additional_data":false},{"leaderboard":"/sota/pansharpening-on-worldview-3-pancollection","task":"Pansharpening","dataset":"PanCollection","model":"DiCNN","rank_in_archive_order":1,"of":2,"metrics":{"ERGAS":"2.7795","Q8":"0.8864","SAM":"3.5170"},"uses_additional_data":false},{"leaderboard":"/sota/pansharpening-on-worldview-3-pancollection","task":"Pansharpening","dataset":"PanCollection","model":"PNN","rank_in_archive_order":2,"of":2,"metrics":{"ERGAS":"2.7756","Q8":"0.8797","SAM":"3.6054"},"uses_additional_data":false},{"leaderboard":"/sota/pansharpening-on-reduced-quickbird","task":"Pansharpening","dataset":"Reduced QuickBird PanCollection","model":"LAGConv","rank_in_archive_order":1,"of":1,"metrics":{"ERGAS":"3.8436","Q4":"0.9314","SAM":"4.5548"},"uses_additional_data":false},{"leaderboard":"/sota/pansharpening-on-reduced-worldview-3","task":"Pansharpening","dataset":"Reduced WorldView-3 PanCollection","model":"LAGConv","rank_in_archive_order":1,"of":1,"metrics":{"ERGAS":"2.3700","Q8":"0.8961","SAM":"3.0414"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1806.08898","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}