Image deblurring and blur kernel estimation of motion blurred colour photographs
| dc.contributor.advisor | Maritz, Milton Frick | en_ZA |
| dc.contributor.author | Engelbrecht, Jan-At | en_ZA |
| dc.contributor.other | Stellenbosch University. Faculty of Science. Department Mathematical Sciences. Applied Mathematics | en_ZA |
| dc.date.accessioned | 2016-03-09T14:57:30Z | |
| dc.date.available | 2016-03-09T14:57:30Z | |
| dc.date.issued | 2016-02-23 | en_ZA |
| dc.description | Thesis (MSc)--Stellenbosch University, 2016 | en_ZA |
| dc.description.abstract | ENGLISH ABSTRACT : Blurred images can be restored using known deconvolution methods. Image blur can be caused by various factors during the image acquisition phase, such as camera defocus or motion. The restoration of blurred images is further complicated by the addition of possible electronic noise. In this report we propose a direct method where the kernel is estimated by nding possible edges in the blurred image using edge detection. A sinogram of the blur kernel is obtained by sampling and then di erentiating across edges in the image for a su cient number of edge orientations. This sinogram is then improved by making use of a sinogram interpolation technique. The application of the inverse Radon transform then yields the kernel which is used to restore the image. | en_ZA |
| dc.description.abstract | AFRIKAANSE OPSOMMING : Vervaagde (d.w.s. uit-fokus) beelde kan herstel word met behulp van bekende dekonvolusiemetodes. Beeldvervaging kan deur verskeie faktore gedurende die verkryging van die beeld veroorsaak word, soos swak fokus of deur beweging van die kamera. Die herstel van vervaagde beelde word verder gekompliseer deur die toevoeging van moontlike elektroniese ruis. In hierdie tesisword 'n direkte metode voorgestel, waar die vervagingskern beraam word deur moontlike rande in die vervaagde beeld met behulp van die randopsporingstegnieke te vind. 'n Sinogram van die vervagingskern word dan verkry deur oor verskeie rande in die beeld te monster vir 'n voldoende aantal randorià ntasies en dan te di erensieer. Die sinogram word dan verbeter deur gebruik te maak van 'n sinogram-interpolasietegniek. Die toepassing van die inverse Radon-transform verskaf die kern, wat gebruik word om die beeld te herstel. | af_ZA |
| dc.format.extent | xiii, 120 pages : illustrations (some colour) | en_ZA |
| dc.identifier.uri | http://hdl.handle.net/10019.1/98753 | |
| dc.language.iso | en_ZA | en_ZA |
| dc.publisher | Stellenbosch : Stellenbosch University | en_ZA |
| dc.rights.holder | Stellenbosch University | en_ZA |
| dc.subject | Blind deconvolution (Applied mathematics) | en_ZA |
| dc.subject | Digital images -- Deconvolution | en_ZA |
| dc.subject | Blur kernel estimation | en_ZA |
| dc.subject | Images, Photographic -- Blurred | en_ZA |
| dc.title | Image deblurring and blur kernel estimation of motion blurred colour photographs | en_ZA |
| dc.type | Thesis | en_ZA |