Detection and tracking of faces in real environments

Abstract

A Stereo Active Vision Interface is introduced which detects frontal faces in real world environments and performs particular active control tasks dependent on changes in the visual field. Firstly, connected skin colour regions in the visual scene are detected by applying a radial scanline algorithm. Secondly, facial features are searched for in the most salient skin colour region while the blob is tracked by the camera system. The facial features are evaluated and, based on the obtained results and the current state of the system, particular actions are performed. The SAVI system is thought of as a smart user interface for teleconferencing, telemedicine, and distance learning. The system is designed as a Perception-Action-Cycle (PAC), processing sensory data of different kinds and qualities. Both the vision module and the head motion control module work at frame rate. Hence, the system is able to react instantaneously to changing conditions in the visual scene.

Face Detection: A Survey

“…Edge features within the head outline are then subjected to feature analysis using shape and position information of the face. More recent examples of edge-based techniques can be found in [9,16,59,115,116] for facial feature extraction and in [32,50,58,60,70,76,108,201,222] for face detection. Edge-based techniques have also been applied to detecting glasses in facial images [79,80].…”

“…The conversion also effectively suppresses the background of other colors and allows the detection of small faces in a natural environment. Other color models applied to face detection include HSV [48,60,81,216], YES [160], YCrCb [2,48,84,130,191,200], YUV [1,123], CIE-xyz [15], L * a * b * [13,109], L * u * v * [63], CSN (a modified rg representation) [90,91] and UCS/Farnsworth (a perceptually uniform color system was proposed by Farnsworth [210]) [208].…”

“…The gazing angle is computed based on the hairline (the border between an individuals hair and skin). Approaches which focus on single camera control include LISTEN [19], SAVI [60], LAFTER [139], Rits Eye [212], and the system of Morimoto and Flickner [135].…”

Appearance and Shape Based Face Recognition Using Backpropagation Learning Neural Network Algorithm with Different Lighting Variations

“…After getting the connected landmark points, the binary image has been taken and treated as shape edge model of the face. The Region Of Interest (ROI) has been chosen according to ROI selection algorithm [26]. Lastly the background noise has been eliminated [27] and appearance based facial feature has been found.…”

Mediated reality using computer graphics hardware for computer vision

“…In [4], a general purpose array of FPGAs was used to achieve face recognition at camera frame rates. While similar hardware could be developed for image warping, such hardware is already incorporated into modern graphics cards.…”