TOPQS color local visual distortion maps
Fragment książki (Rozdział w książce)
| Status: | |
| Autorzy: | Skublewska-Paszkowska Maria, Smołka Jakub |
| Wersja dokumentu: | Drukowana |
| Arkusze wydawnicze: | 0.5 |
| Język: | angielski |
| Strony: | 277 - 284 |
| Efekt badań statutowych | NIE |
| Materiał konferencyjny: | NIE |
| Publikacja OA: | NIE |
| Abstrakty: | polski |
| Image quality is an important aspect which should be taken into consideration during image processing. Quality can be measured using various methods, one of which is a perceptual Picture Quality Scale measure. It allows to verify whether the distortions in the image are visible to human beings and to what extent. Unfortunately this measure needs assessments from a group of observers. On their basis a single value is computed that corresponds to the amount of visual distortions present in image. Combining this measure together with a neural network allows to eliminate the need for human observers. This simplifies the assessment of image deformations and permits implementation that can be used for visualization of the distortions. For the purpose of the following paper the PQS measure was implemented with neural network resulting in a new measure called TOPQS. It was in turn used to obtain color local image visual distortions. They were computed using two images: the original and the processed one. Corresponding parts of these images were compared and the TOPQS value, which evaluates local distortions, was obtained. Once the images were analyzed local distortion maps were generated. |
