“…Reconstructing real ground truth images with the proposed super-resolution approach At this stage, the performance of the proposed method was compared quantitatively and qualitatively to state-of-the-art and conventional techniques, such as bicubic interpolation, FIGURE 6 Comparing the image quality of the proposed method to that of existing algorithms for a sample image of chest X-ray datasets at an x2 magnification scale dictionary-based image enhancement methods (FL [31], A+ [32]), self-similarity-based method Self-ExSR [33], and artificial neural network-based techniques (LapSRN [34], Ms-LapSRN [35], DRRN [36], FSRCNN [33], SRCNN [6], SCN [37], DRCN [38], VDSR [38], RDN [39], LapSRN [34], EDSR [40], CARN [41], MemNet [42], RCAN+ [43], SRCLIQUENET+ [44]), for magnification factors of 2, 3, and 4. Furthermore, SET5, SET14, and BSDS100 databases, which contained images of natural scenes, as well as the URBAN100 dataset, which contained images of architectural challenges, were evaluated.…”