Improved Error Reduction and Hybrid Input Output Algorithms for Phase Retrieval by including a Sparse Dictionary Learning-Based Inpainting Method

<div>An example of applying the KSVD method to restore the damaged image. (a) Original image of the painting “Girl with a Pearl Earring” by Johannes Vermeer. (b) Damaged image containing 25% missing pixels has PSNR = 13.77 dB. (c) Reconstruction result using the KSVD method has PSNR = 31.56 dB. (d) The updated/learned dictionary. Each patch represents a column vector or an atom of the dictionary matrix <span class="nowrap"><svg height="8.79011pt" id="M78" style="vertical-align:-0.04980946pt" version="1.1" viewbox="-0.0498162 -8.7403 10.0297 8.79011" width="10.0297pt" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"><g transform="matrix(.013,0,0,-0.013,0,0)"><path d="M41 650V615C111 608 121 602 121 516V134C121 48 111 42 32 35V0H289C413 0 518 27 592 83C668 141 717 229 717 351C717 446 678 517 628 564C564 625 467 650 341 650H41ZM255 540C255 569 257 586 265 595C272 602 292 609 328 609C397 609 450 589 491 547C537 502 566 426 566 331C566 152 474 43 340 43C270 43 255 68 255 140V540Z"></path></g></svg>.</span></div>

International Journal of Optics

fig4

Figure 4

Figure 4: Improved Error Reduction and Hybrid Input Output Algorithms for Phase Retrieval by including a Sparse Dictionary Learning-Based Inpainting Method 

Figure 4 | Improved Error Reduction and Hybrid Input Output Algorithms for Phase Retrieval by including a Sparse Dictionary Learning-Based Inpainting Method 