{"title":"Image Quality Improvement of Surveillance Camera Images by Learning-based Denoising Method Utilizing Noise2Noise","authors":"Akira Kuchida, T. Goto","doi":"10.1145/3576938.3576943","DOIUrl":"https://doi.org/10.1145/3576938.3576943","url":null,"abstract":"In recent years, the number of surveillance cameras installed has increased. Surveillance cameras need to be able to capture images even under poor shooting conditions such as low exposure. However, noise may be generated in the captured images under such environments. Although there have been many studies on image denoising, most of them target only synthetic noise such as Gaussian noise or real image noise such as the SIDD dataset and have not demonstrated sufficient performance for captured images. In this paper, we investigate the construction of an effective CNN model for real image noise using Noise2Noise. In addition, Noise2Noise has the problem of significantly degraded performance compared to normal learning when data is small. Therefore, we propose a learning method that can build models with good performance even when data is small, by pre-training with an open dataset such as SIDD and then re-training with Noise2Noise.","PeriodicalId":191094,"journal":{"name":"Proceedings of the 2022 5th International Conference on Digital Medicine and Image Processing","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126989489","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Reconstruction From Limited-Angle Projections Based on a Transformation","authors":"M. Hjouj, Muntaser S. Ahmad","doi":"10.1145/3576938.3576942","DOIUrl":"https://doi.org/10.1145/3576938.3576942","url":null,"abstract":"ABSTRACT Limited angle computed tomography (CT) is common in medical and industrial applications where incomplete projection data can cause artifacts in the reconstructed image. we propose a new algorithm for reconstructing a density function , (an image), in the plane from a limited number of Radon projections on a range of angles for some acute angle Assuming that is subjected to a linear transformation to produce a new image ; the new image is then recovered. In fact, the well-known relation between the desired image and the transformed image, in the transform domain, is used. This relation allows us to map a uniformly distributed view angles over to some range of view angles that are available for the original image. It is known that uniformizing the view angles of projections improves the performance of some algorithms such as the filter back projection algorithm. In this way, the transformed image is reconstructed from the uniformly distributed angles that correspond to a limited number of projections. The effectiveness of the proposed approach is validated by simulation and by applications to synthetic images.","PeriodicalId":191094,"journal":{"name":"Proceedings of the 2022 5th International Conference on Digital Medicine and Image Processing","volume":"124 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115612006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}