Facial Image Enhancement Limitation of Using Spatial Domain Histogram Enhancement Methodologies

This study evaluates various spatial domain image enhancement procedures, as well as facial image enhancement methods based on histogram improvement. For colour pictures, RGB space is separately upgraded and mixed to reproduce the outcomes. For the performance evaluation, the paper evaluated histogram equalization (HE), Bi histogram equalization (BBHE), and CLAHE methods. The study discusses the limitations of spatial domain approaches. For the performance measurements, the YouTube makeup face pictures database and the Yale Face database are used. The spatial case of facial makeup database is considered as level of difficulties for enhancement quality. The entropy and SNR analysis is carried out. Despite that, every SDE is the most colour-sensitive. CLAHE outperforms with 30.69 dB SNR and mean brightness of 173.557. Although good contrast improvement is provided, it is determined that spatial domain approaches are sensitive to brightness variations and so are unsuitable for face picture enhancement even after brightness preservation.