Robust license-plate recognition method for passing vehicles under outside environment

作者：Tomoki Naito, T. Tsukada, Koki Yamada, Kazuhiro Kozuka, Shin Yamamoto · 发表于：IEEE Transactions on Vehicular Technology · 年份：2000 · DOI：10.1109/25.901900 · 被引用次数：224 · 研究领域：Vehicle License Plate Recognition、Advanced Neural Network Applications、Image and Object Detection Techniques

Novel methods to recognize license plates robustly are presented. A sensing system with a wide dynamic range has been developed to acquire fine images of vehicles under varied illumination conditions. The developed sensing system can expand the dynamic range of the image by combining a pair of images taken under different exposure conditions. In order to avert blurring of images against fast passing vehicles, a prism beam splitter installed a multilayered filter, and two charge-coupled devices are utilized to capture those images simultaneously. Furthermore, to extend the flexibility of camera placement, a recognition algorithm that can be applied to inclined plates has been developed. The performance of recognizing registration numbers on license plates has been investigated on real images of about 1000 vehicles captured under various illumination conditions. Recognition rates of over 99% (conventional plates) and over 97% (highly inclined plates) showed that the developed system is quite effective for license-plate recognition.