Edge Map Extraction of High-Resolution Facial Images

This work focuses on edge detection, a vital task in several fields, such as computer vision, image processing, and pattern recognition. Since exact facial analysis and recognition have numerous applications in areas like biometrics, surveillance, and human-computer interaction, this study’s focus is specifically on obtaining precise edge maps of facial pictures with exceptional resolution. Several essential processes are carried out in order to produce accurate edge maps. In order to lessen noise and attenuate the very low-level features in high-resolution photographs, a smoothing operation is first conducted. Utilizing a variety of edge detection methods, such as directional gradients and cumulative gradient magnitude, variations in pixel intensities along various directions are examined. The Zhang-Suen technique, renowned for its efficiency in thinning binary pictures, is used to remove extraneous pixels from the discovered edges to further refine them. By maintaining only the necessary boundary information, this phase ensures that the edge maps produced are more accurate. According to subjective and comparative evaluations and the outcomes of applying the algorithm to face photos from various datasets, the suggested approach is suitable for building edge maps of facial photographs with high resolution as well as low resolution.