RKMajji.ppt

Hand geometry uses measurements of hands for identification. It captures images of hands and extracts measurements like finger lengths and widths. Preprocessing steps include binarization, rotation, and contour extraction. Features are extracted and classified using models like Gaussian mixture models. Researchers have achieved hit rates up to 88.89% using length and width features in the first group, and fingertips in the second group. Observations note GMM obtains best results while requiring more memory than other comparison methods. The conclusion suggests using hand geometry for medium security and combining it with other biometrics like palm prints.