Fast Evolutionary Algorithm based Identifying Surgically Distorted Face for Surveillance Application
R. Mallika Pandeeswari, K. Deepthyka, M. Abinaya, V. Deepa, R. Kabilan, J. Glorintha · 2022 International Conference on Sustainable Computing and Data Communication Systems (ICSCDS) · 2022
This research, which employs granular computing and hybrid spatial feature extraction, presents face identification with surgically changed characteristics. The increased popularity of cosmetic surgery and its influence on automatic face recognition has sparked the curiosity of researchers. The face detection module generates face pictures with normalised intensity, homogenous size and shape, and just the face region. The Difference of Gaussian (DOG) structures will then be generated from successive iterations of Gaussian photographs. Granulation divides face pictures into different resolutions and provides blurriness, edge information, and noise in a face image. The Gabor filter bank is then used to extract characteristics from face regions in order to differentiate between illumination variations. These characteristics are merged in order to reliably identify a huge samples, and they will be matched with stored real face samples to identify them. The simulated results of detecting surgically changed face photographs show that utilising granulation and hybrid spatial and spectral descriptors has greater selectivity and recognition accuracy.