Face detection systems are used in different multimedia applications for the facial expression recognition

Face detection systems are used in different multimedia applications for the facial expression recognition, facial feature selection and extraction, and face identification using websites and Application, and automatic face recognition from digital images. Without face detection process and facial feature extraction machine cannot recognize a human face. Various face detection methods and algorithms are used to detect the frontal-faces from digital images. Detection of a non-frontal face (i.e., face with different rotations) is complex task. To overcome this problem, we proposed a Rotation Invariant Face Detection System (RIFDS) for fast and accurate detection of faces at the different angle of rotations. The advantage of proposed face detector is that it can detect the face from the digital image at different rotations (i.e., ±30?, ±45?, ±60?, ±90?, ±120?, ±135?, ±150?, 180?, ±210?, ±225?, ±240?, ±270?, ±300?, ±315?, ±330? and 360?). We also compared the proposed system with other algorithms used for face detection. The proposed face detector is tested on JAFFF Japanese Database contain 215 images of the female with different facial expression and we achieve 100% face detection accuracy at different rotations. Proposed face detection system achieves 90% accuracy on ORL database, 84.76% accuracy on CMU database, 92.3% accuracy on LFW database and 90% accuracy on MIT-CMU database. The developed face detection system is able to detect face at different image resolutions (e.g., 84×60, 128×120, 105×105, 110×110, 106×149, 92×112, 82×82, 115×115, 256×256, 512×512, 32×32, 64×64 and 250×250). The face detection speed of proposed system is 0.045154 seconds.