| _version_ | 1866901923527917568 |
|---|---|
| author | Austin Siby, Ruth Nhanu Gawade, Suhas Rautmare |
| author_facet | Austin Siby, Ruth Nhanu Gawade, Suhas Rautmare |
| contents | <div> <div>Image compression has become essential for managing visual data in modern digital systems [1].Every photograph uploaded, video streamed, or medical scan stored relies on compression to reduce file sizes, enabling faster transmission and efficient storage utilization. As imaging devices produce increasingly high-resolution content and connected systems proliferate, visual data volumes grow exponentially [23]. Storage infrastructure faces capacity constraints, network bandwidth proves inadequate, and efficient data management becomes imperative. Compression addresses these challenges but introduces quality degradation through information loss [13]. File size reduction necessitates data discarding, manifesting as blurred regions, blocking artifacts, lost detail, and color shifts. While acceptable for casual applications, critical domains like medical imaging and satellite monitoring require higher fidelity where minor distortions compromise decision-making [20], [21]. This study quantitatively evaluates compression-induced information loss using established metrics rather than subjective assessment. Original images are systematically compared against compressed versions using Mean Squared Error (MSE), Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity Index (SSIM) [2], [10]. Multiple compression techniques are tested across varying quality levels, revealing performance characteristics under increasing compression pressure. The analysis provides empirical evidence regarding storage efficiency and quality preservation trade-offs. By mapping these patterns through objective measurement, this research delivers data-driven guidance for selecting compression settings aligned with application requirements.</div> </div> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_18719892 |
| institution | Zenodo |
| language | |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Analyzing Image Compression Through Quantitative Analysis Austin Siby, Ruth Nhanu Gawade, Suhas Rautmare <div> <div>Image compression has become essential for managing visual data in modern digital systems [1].Every photograph uploaded, video streamed, or medical scan stored relies on compression to reduce file sizes, enabling faster transmission and efficient storage utilization. As imaging devices produce increasingly high-resolution content and connected systems proliferate, visual data volumes grow exponentially [23]. Storage infrastructure faces capacity constraints, network bandwidth proves inadequate, and efficient data management becomes imperative. Compression addresses these challenges but introduces quality degradation through information loss [13]. File size reduction necessitates data discarding, manifesting as blurred regions, blocking artifacts, lost detail, and color shifts. While acceptable for casual applications, critical domains like medical imaging and satellite monitoring require higher fidelity where minor distortions compromise decision-making [20], [21]. This study quantitatively evaluates compression-induced information loss using established metrics rather than subjective assessment. Original images are systematically compared against compressed versions using Mean Squared Error (MSE), Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity Index (SSIM) [2], [10]. Multiple compression techniques are tested across varying quality levels, revealing performance characteristics under increasing compression pressure. The analysis provides empirical evidence regarding storage efficiency and quality preservation trade-offs. By mapping these patterns through objective measurement, this research delivers data-driven guidance for selecting compression settings aligned with application requirements.</div> </div> |
| title | Analyzing Image Compression Through Quantitative Analysis |
| url | https://doi.org/10.5281/zenodo.18719892 |