|
|
Effect of Noise on Image Compression
Pavlík, Jiří ; Svoboda, Pavel (referee) ; Bařina, David (advisor)
This thesis examines effect of different types of noise on performance of significant image compression formats. Three different types of noise are examined, Gaussian noise, Gaussian noise with bigger granularity and shot noise. Two older loosy image compression formats JPEG and JPEG 2000 are compared with newer format WebP. This examination is based on results of experiments with images, which are intentionally degraded by noise, then compressed to examined formats and compared with result of experiments for original images without noise. The effect of noise on performance of comparised formats is based on image quality metric SSIM. From results of experiments, Gaussian noise with bigger granularity seems to be the least distruptive from all types of noises. The other two types of noise have much more dwindling effect on quality of images.
|
|
|
Dithering the image
Štefaník, Pavel ; Zátyik, Ján (referee) ; Rajmic, Pavel (advisor)
Bachelor work deals with the reducing of picture color depth by various dithering methods. These methods are compared between each other and demonstrated on various types of pictures. Each method is described and implemented in MATLAB. The results are compared from different angles: time duration of algorithm, survey based on form, my opinion and SSIM comparison.
|
|
|
Comparison of Lossy Image Compression Formats
Koláček, Ladislav ; Svoboda, Pavel (referee) ; Bařina, David (advisor)
This bachelor thesis deals with comparing of lossy image compression formats. The chapters describe principles of image compression and lossy formats themselves. Specifically the lossy formats JPEG, JPEG 2000, JPEG XR and WebP. The thesis includes theoretical comparison of the formats mentioned above, along with the methods used for measuring image quality. The practical part contains the results of the quality measurements of each format and description of design together with implementation of the tutorial.
|
|
|
Dithering the image
Štefaník, Pavel ; Zátyik, Ján (referee) ; Rajmic, Pavel (advisor)
Bachelor work deals with the reducing of picture color depth by various dithering methods. These methods are compared between each other and demonstrated on various types of pictures. Each method is described and implemented in MATLAB. The results are compared from different angles: time duration of algorithm, survey based on form, my opinion and SSIM comparison.
|
|
|
Comparison of Lossy Image Compression Formats
Koláček, Ladislav ; Svoboda, Pavel (referee) ; Bařina, David (advisor)
This bachelor thesis deals with comparing of lossy image compression formats. The chapters describe principles of image compression and lossy formats themselves. Specifically the lossy formats JPEG, JPEG 2000, JPEG XR and WebP. The thesis includes theoretical comparison of the formats mentioned above, along with the methods used for measuring image quality. The practical part contains the results of the quality measurements of each format and description of design together with implementation of the tutorial.
|
|
|
Effect of Noise on Image Compression
Pavlík, Jiří ; Svoboda, Pavel (referee) ; Bařina, David (advisor)
This thesis examines effect of different types of noise on performance of significant image compression formats. Three different types of noise are examined, Gaussian noise, Gaussian noise with bigger granularity and shot noise. Two older loosy image compression formats JPEG and JPEG 2000 are compared with newer format WebP. This examination is based on results of experiments with images, which are intentionally degraded by noise, then compressed to examined formats and compared with result of experiments for original images without noise. The effect of noise on performance of comparised formats is based on image quality metric SSIM. From results of experiments, Gaussian noise with bigger granularity seems to be the least distruptive from all types of noises. The other two types of noise have much more dwindling effect on quality of images.
|
guest ::
