{"title":"An Improved Image Super-Resolution Reconstruction Method Based On LapSRN","authors":"Lei Kong, L. Jiao, Feng Jia, Kai Sun","doi":"10.1109/CISP-BMEI53629.2021.9624332","DOIUrl":null,"url":null,"abstract":"With the gradual maturity of the traditional static image recognition field, super-resolution reconstruction based on deep neural networks is a research hotspot and difficulty in the field of computer vision. In particular, most single-frame image super-resolution methods have problems such as loss of high-frequency information, noise introduced in the up-sampling process, and difficulty in determining the interdependence between each channel of the feature map when reconstructing the predicted image. In order to solve the above problems, we introduce back projection mechanism into the LapSRN network in this paper. By introducing the back projection mechanism effectively improved the consistency between the extracted image feature data and the target feature data feature, and thereby improved the reconstructed image parameters. Experiments show that the improved network can achieve better performance than LapSRN.","PeriodicalId":131256,"journal":{"name":"2021 14th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2021 14th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/CISP-BMEI53629.2021.9624332","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
With the gradual maturity of the traditional static image recognition field, super-resolution reconstruction based on deep neural networks is a research hotspot and difficulty in the field of computer vision. In particular, most single-frame image super-resolution methods have problems such as loss of high-frequency information, noise introduced in the up-sampling process, and difficulty in determining the interdependence between each channel of the feature map when reconstructing the predicted image. In order to solve the above problems, we introduce back projection mechanism into the LapSRN network in this paper. By introducing the back projection mechanism effectively improved the consistency between the extracted image feature data and the target feature data feature, and thereby improved the reconstructed image parameters. Experiments show that the improved network can achieve better performance than LapSRN.