Poisson Noise Removal Using Non-convex Total Generalized Variation

Xinwu Liu,Yingying Li
DOI: https://doi.org/10.1007/s40995-021-01203-3
2021-08-19
Abstract:<p class="a-plus-plus">As is well-known that the total generalized variation model performs well in reducing the staircasing effect while removing noise, but it tends to cause the undesirable edge details blurring. To overcome this drawback, the current paper introduces the non-convex restriction into the total generalized variation regularizer and constructs an improved edge-preserving optimization model for Poissonian images restoration. For solving the minimization problem, we propose an efficient alternating minimization method by skillfully combining the classical iteratively reweighted <span class="a-plus-plus inline-equation id-i-eq1"><span class="a-plus-plus equation-source format-t-e-x"><span class="mjpage"><svg xmlns:xlink="http://www.w3.org/1999/xlink" width="2.024ex" height="2.509ex" style="vertical-align: -0.671ex;" viewBox="0 -791.3 871.4 1080.4" role="img" focusable="false" xmlns="http://www.w3.org/2000/svg"><g stroke="currentColor" fill="currentColor" stroke-width="0" transform="matrix(1 0 0 -1 0 0)"> <use xlink:href="#MJMAIN-2113" x="0" y="0"></use> <use transform="scale(0.707)" xlink:href="#MJMAIN-31" x="590" y="-213"></use></g></svg></span></span></span> algorithm and primal-dual framework. Some visual experiments presented in the illustration section, which are compared with some related denoising methods, demonstrate the better performance of the developed scheme in staircase artifacts reduction and image features protection. Besides, the measurable comparisons also indicate that our outcomes enjoy the best restoration accuracy against other popular competitors.</p><svg xmlns="http://www.w3.org/2000/svg" style="display: none;"><defs id="MathJax_SVG_glyphs"><path stroke-width="1" id="MJMAIN-2113" d="M345 104T349 104T361 95T369 80T352 59Q268 -20 206 -20Q170 -20 146 3T113 53T99 104L94 129Q94 130 79 116T48 86T28 70Q22 70 15 79T7 94Q7 98 12 103T58 147L91 179V185Q91 186 91 191T92 200Q92 282 128 400T223 612T336 705Q397 705 397 636V627Q397 453 194 233Q185 223 180 218T174 211T171 208T165 201L163 186Q159 142 159 123Q159 17 208 17Q228 17 253 30T293 56T335 94Q345 104 349 104ZM360 634Q360 655 354 661T336 668Q328 668 322 666T302 645T272 592Q252 547 229 467T192 330L179 273Q179 272 186 280T204 300T221 322Q327 453 355 590Q360 612 360 634Z"></path><path stroke-width="1" id="MJMAIN-31" d="M213 578L200 573Q186 568 160 563T102 556H83V602H102Q149 604 189 617T245 641T273 663Q275 666 285 666Q294 666 302 660V361L303 61Q310 54 315 52T339 48T401 46H427V0H416Q395 3 257 3Q121 3 100 0H88V46H114Q136 46 152 46T177 47T193 50T201 52T207 57T213 61V578Z"></path></defs></svg>
What problem does this paper attempt to address?