Authors : Mohammad Al-Khmess and Mohmad- Bassam Kurdy
Abstract: In this research, we proposed a system to automatic facial recognition photo tagging. Automatic tagging in photo is challenging of both side the face detection and the face recognition standpoints we need face detection and then recognition. We rely on the Color space YCrCb as well as the Color space HSV to extract skin region. Then we examined skin face candidate by searching for the eyes using a Viola Jones algorithm. When we detect an eye, this region will be a face. Regions that do not have eyes are dismissed because they are likely to be either legs or hands. Later, we divide faces to tag it then save them as a database or used as faces of query to be tagged. Eigenface algorithm does the process of recognition. Using two space color improves extracting skin. The Viola Jones algorithm ensures the accuracy of eyes searching results. But it requires long processing time. In this research the search performance of the algorithm is improved to achieve quick results. Therefore, result has been very good when tagging the same photo.
Mohammad Al-Khmess and Mohmad- Bassam Kurdy, 2020. Automatic Facial Recognition Photo Tagging using Two Spaces Color and Eigenface Algorithm. International Journal of Soft Computing, 15: 7-12.