2025-03-102025-03-102022-12-15SILVA, Iury Glabson Oliveira. Avaliação experimental da compressão de vídeos de estruturas civis utilizando técnicas de análise modal. Orientador: Reginaldo Cordeiro dos Santos Filho. 2022. 48 f. Trabalho de Conclusão de Curso (Bacharelado em Ciência da Computação) – Faculdade de Computação, Instituto de Ciências Exatas e Naturais, Universidade Federal do Pará, Belém, 2022. Disponível em:. Acesso em:.https://bdm.ufpa.br/jspui/handle/prefix/7808In the area of structural health monitoring, modal analysis is used to verify the physical condition of structures, by extracting relevant features to damage identification. To achive this, modal properties (natural frequencies, vibration modes, damping ratios) are generally utilized, because changes in the structures physical characteristics can be seen in it’s modal properties. Furthermore, these properties also represent the brute data, wich can be videos, sensor measurements, and others, in a reduced form. In this work, we explore the potential of modal properties, extracted during modal analysis, as compressed versions of the original measurements, specifically in videos with vibrating structures. To achive this, a framework that extract modal properties from videos with vibrating structures by applying dimension reduction and feature extraction techniques, like principal component analysis and blind source separation, is implemented and adapted to export the result of its procedures (modal coordinates, static background and complex vibration modes). Furthermore, some of the most used codecs in the literature, H.264/AVC, H.265/HEVC, VP9 and AV1, were selected and implemented in a pipeline to gather data that could be compared with those collected from the developed framework. With that in mind, some video quality metrics, like PSNR, MS-SSIM and VMAF, were selected, those being essential to collect the data that will be analyzed. After implementing those codecs, metrics and the modal analysis framework, a comparative analysis is executed to achive this paper principal objective. The collected data shows that, in fact, modal analysis do have the potential to compress video measurements. The results describe that, for the same compression rate (all implementations where analysed in the same compression rate), the quality of videos compressed using the framework are equivalent to those compressed with the selected codecs of the literature in metrics that uses the Human Vision System (HVS) as a decisive factor in the video quality perception, as MS-SSIM and VMAF. And for those same metrics, the results of the framework were consistently better than those of the codec H.254/AVC.Acesso AbertoAnálise modalCompressão de vídeosStructural health monitoringExtração de característicasModal analysisVideo compressionStructural health monitoringFeature extractionCNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAOTrabalho de Curso - Graduação - Monografia