Adaptive Detecting the Time of the Object Appearance in a Sequence of Random Images

G.I. Salov (Russia)


control problem, detecting problem, learning algorithm, sequence of images, optimal stopping time.


Two new methods are proposed for the solution of an im portant class of multidimensional detection problems. The first is extended method developed by the author for adap tive detection of a signal hidden in nose to adaptive detec tion of an object. This method is based on the approxima tion for the (unknown) likelihood functional and the learn ing procedure. The second is application the first method to the problem of quickest detecting the time of the appear ance of the object in sequence of independent random im ages. This method is based on the properties of a sequence of cumulative sums.

Important Links:

Go Back