Pattern recognition is derived from the will to imitate the sense of the creatures

object searching in an image aims at imitating the human' s visual ability. In this paper

utilizing the results of the vision neurophysiology and the psychology of cognizance

and aiming at the specific pattern location in the natural environment

we put forward a Bionic Object Exploring Strategy——BOES. In a license plate recognition system

an algorithm of succinctness and practicality on this strategy has developed to solve the difficulty of locating license plates rapidly and precisely. In our algorithm

to deal with the complex

uncontrollable background of the natural environment and the variability of the color

brightness and size

we extract the license plate pattern' s most permanent characters

design a set of filters to mask the background

organize the image data in a series of steps by spatial and frequencial filtering

and divide the task into several layers. This system can locate the license plate in an image of complex environment within 0.5 second on a PentiumⅡ 300Hz machine

and the rate of success is over 95%.