近年來隨著車牌影像辨識系統技術的進步，實際的潛在應用也日益增加，例如查緝贓車、取締違規車輛以及無人化停車場自動收費系統等。然而在實際上，車牌辨識系統目前尚無法便捷、快速地普及化，其主要原因之一是在辨識技術上仍有需要改善的地方，如車牌定位率、字元辨識率等。因此本研究針對此部份期望能提升車牌影像辨識系統辨識的準確率。 在車牌辨識系統中主要分為車牌定位以及車牌辨識兩部分，其中在車牌辨識部分中選取車牌二值化的門檻值對車牌字元分割以及辨識中扮演著很重要的因素，門檻值若選的不好將會造成字元切割困難和辨識錯誤等問題發生，因此本文加入照度參數以經驗法則作為基礎，利用大量車牌資料去做門檻值的篩選和調整，經由實驗的結果來選擇其最佳的數值作為門檻值範圍。 最後經由實驗證明，利用本論文提出的二值化門檻值範圍不管是在燈光明亮的地方還是昏暗之處，都能夠大幅的提升字元切割和車牌辨識的成功率，使得整個車牌辨識系統可以發揮出更好的效能。 Over recent years, as the vehicle license plate recognition system technology continues to improve, the actual potential applications are also on the rise. Common application examples are stolen cars identification, traffic violations ticketing, and the toll collecting systems of man-free parking lots. Yet in practical implementation, a major reason that the license plate recognition system is unable to rapidly and expeditiously become prevalent lies in how the recognition technology leaves room to be improved upon, such as the license plate positioning ratio, character recognition ratio and so on. As a result, this study intends to focus on this area such that the recognition accuracy of the license plate recognition system can be further improved. License plate recognition system is divided into two parts as the license plate positioning and license plate recognition. The binarization of threshold plays an important role in the license plate character segmentation and recognition which used in the image edge detection. Unsuitable and inappropriate threshold will cause character cutting difficulties and recognition errors. This paper will add illumination parameters and use a large number of license plate information to adjust the threshold value, and then select the best value for threshold according to empirical result. Finally, experiments show that the derived threshold values will significantly enhance the success rate of character cutting and license plate recognition whether it is in bright light or a dark place, and also makes the license plate recognition system perform better.