2-d Object Recognition Using Surveillance Video Processing on Daubechies Wavelet Decomposition

The project suggests effective motion recognition built on background subtraction using dynamic threshold approach with mathematical morphology. Here the techniques frame differencing, dynamic threshold based detection will be used. In addition to dynamic threshold, mathematical morphology is also used within ability of greatly attenuating color dissimilarities has generated by background motions while still highlighting moving objects. After the object foreground detection, the parameters like velocity motion, speed will be determined. Finally the simulated results will be shown that used background subtraction with daubechies wavelet decomposition approach is effective rather than previous background subtraction methods.