Low-light robust face super resolution via morphological transformation based locality-constrained representation. (March 2023)

• Record Type: Journal Article
• Authors: Kumar, Banti; Rajput, Shyam Singh
• Is Part Of: Computers & electrical engineering. Volume 106(2023)
• Publication Date: 2023-03
• Subjects: Low-light -- Morphological transformation -- Super-resolution -- Hallucination; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2023.108612
• Languages: English
• ISSNs: 0045-7906

Abstract

Abstract: A new low-light robust face super-resolution (SR) model via morphological transformation-based locality-constrained representation (MTLcR) is proposed to improve the robustness of the current SR algorithms against low-light problems. The proposed model first tries to detect the amount of low-light effects present in the input images with the help of dictionary images of the same domain. Further, the same amount of effect is applied to low-resolution (LR) dictionary images with the purpose of minimizing the reconstruction error while representing the input image as a weighted linear combination of LR dictionary images. In addition to this, morphological transformation is introduced in the objective function to further improve the reconstruction weights. The simulation results on a publicly available data set and some real-world images suggest the better reconstruction capability of the proposed model under low-light scenario as compared to other state-of-the-art works.