آمار
| تعداد نشریات | 50 |
| تعداد شماره‌ها | 2,211 |
| تعداد مقالات | 20,279 |
| تعداد مشاهده مقاله | 24,205,053 |
| تعداد دریافت فایل اصل مقاله | 22,094,192 |
ارائه مدل بهینه ریاضی مبتنی بر نمایش تُنُک در جهت بهبود بازسازی تصاویر | ||
| Journal of New Researches in Mathematics | ||
| مقالات آماده انتشار، پذیرفته شده، انتشار آنلاین از تاریخ 08 مهر 1402 | ||
| نوع مقاله: research paper | ||
| شناسه دیجیتال (DOI): 10.30495/jnrm.2023.73817.2426 | ||
| نویسنده | ||
محمّدسعید علمداری  
| ||
| گروه ریاضی کاربردی، دانشکده ریاضی، دانشگاه خواجه نصیر الدین طوسی، تهران، ایران | ||
| چکیده | ||
| In recent years, signal and image processing with the help of sparse reconstruction models has been highly regarded by researchers and has been successful in a wide range of industrial, medical, military, security systems, etc. applications. Providing a simple and compact description of the signal in terms of a linear combination of a small number of elements of a set called a dictionary consisting of basic signals is the basic idea of sparse reconstruction models. One of the basic issues in these models is the reconstruction of the sparse signal from its noisy version, which is very challenging. The aim of this article is to present an optimal mathematical model based on sparse representation in order to improve image reconstruction, which uses smoothed L0 norm instead of L0 norm and transforms the quadratic subproblem into an optimization problem with equality constraints and transfers the inequality constraint to the objective function. As a result, it creates a new approach to solve sub-problems in less time based on estimating a more accurate solution and with a higher convergence speed. Examining the experiments shows that the proposed method has a very favorable performance compared to sparse image reconstruction algorithms and has a high success rate in image reconstruction. | ||
| کلیدواژه‌ها | ||
| Sparse Representation؛ Modeling؛ Image Reconstruction؛ Optimization؛ Dictionary Learning | ||
|
آمار تعداد مشاهده مقاله: 42 |
||
