Browsing by Author "SYED FAWAD RAZA RIZVI"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Modified borman
    (UMT.Lahore, 2019) SYED FAWAD RAZA RIZVI
    Facial Feature/Fiducial point detection is one of the most useful topics of these days in computer vision research. The need of this type of detections is increasing in every society. The most highlighted area which required FFPD (Facial Feature Point Detection) is security precautions. Because identifying between normal and abnormal people (in this case abnormal means thieves, terrorists etc.) is a big task for security agencies of any country through surveillance. Now a bit of introduction about our modified algorithm. Implementation have been done by developing Modified BoRMaN for facial feature points detection from video sequences, datasets consist of random moments, datasets with illumination changes etc. With this improved algorithm identification of faces and detection of facial fiducial points could be done easily and accurately. Our proposed Modified BoRMaN is an extension of Temporal BoRMaN which uses more points with good accuracy. Moreover, we have tested this proposed model for random motion and video sequences as well. The existing algorithm which is used for comparison in this research is BoRMaN. The goal of this research is to improve the accuracy FFP’s (Facial Feature Points) and detect points from moving face as well. This temporal algorithm will be helpful in security precautions as well.