Answers are grounded in this paper and cite their sources.
This work is a review of the experience in solving the problems of retouching and reconstructing the content of continuous lacunas in images. Restoring continuous (long) gaps is significantly more difficult than doing similar work with sparse data, since prolonged loss leads to the loss of signal phase and dynamics. The main problem is that, given the significant size of the lacunas, most algorithms cannot maintain sufficient continuity and consistency of the signal or image structure. It has been shown that image inpainting methods or neural networks work well when retouching or masking the content of a gap. And if it is necessary to restore the contents of the lacunas, it is preferable to use methods that operate in the frequency domain (for example, the interpolation method of the sequentially calculated Fourier spectrum).
No takes yet. Share an insight, caveat, or question.
Alexander V. Kokoshkin (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: