The algorithms detection objects of video surveillance on statistical pattern

Some problems of pattern recognition are solving with the Bayes likelihood ratio test. To construct the likelihood ratio, we must have the conditional probability density function for each class. Even if we can obtain the densities, the likelihood ratio test may be diffi cult to implement; time and storage requirements for the classifi cation process may be excessive. But if the conditional probability density is not defi ned by a formula, and is defi ned as a table of values, then it is solving the diffi cult to implement.