Method of License Plate Location Based on Corner Feature

License plate location (LPL) plays an important role in the license plate recognition (LPR). The license plate characters contain abundant corners which distribute intensively and regularly. Accord to the feature, a novel method to locate the license plate accurately and fast is proposed in this paper. The method is dived into three steps which are corners detection, finding out candidate license plates and locating the license plate accurately. In the course of locating license plate, three thresholds are used. First, the reasonable threshold is used to eliminate a great number of non-corner pixels to speedup corner-detections. Second, a big threshold is selected with a slid window to find out candidates. Finally, the region of candidate license plates are expanded, all the corners in it were stand out with a smaller threshold, and the license plate is located accurately using clustering analyses. Experiments have been conducted, 1200 license plate images are tested, the accuracy is 98.42% and the average consuming time is 20ms