License-plate location and character segmentation are two key components of the automatic vehicle license recognition system. An edge-texture-color based algorithm to locate the plate region and a template matching based algorithm to segment characters are proposed. Using these algorithms and a character recognition kernel, an automatic vehicle license recognition system was finished. The performance of the system was investigated on many real car images with complex background captured under various illumination conditions. The experimental results show that the proposed algorithms have high accuracy and robustness.