- 期刊
應用影像分析技術於停車場之進出管理
Entering-and-Exiting Management of Parking Lots by an Application of Image-Analysis Techniques
摘要
國人擁有自用小客車的比例逐年增多，停車場的增加速度卻趕不上車輛的成長，停車問題愈顯嚴重。目前停車場對出入車輛的管理方式，是在出入口位置，對車輛進行定點式的收費管理。這種管理方式，很容易在出入口造成車輛排隊等候的狀況。另外，停車場往往有不同等級客戶的車輛出入，希望可以利用不同的進出方式，來滿足各層級客戶對進出速度、隱私保護等的不同要求。因此，停車場進出的自動化管理是亟待解決的問題。車輛牌照如同個人的身份證一樣具有唯一性，如能運用電腦的快速處理能力於車牌影像辨識處理上，並規劃不同層級客戶的特有進出方式，將可解決目前停車場進出在管理上的諸多問題。因此，本論文提出一套機制，藉由影像分析技術及不同進出方式的規劃，有效縮短不同權限車輛進出停車場的時間，以提升停車場進出的管理效率。
並列摘要
In Taiwan the percentage of persons owning private automobiles is increasing annually. However, the growth in the number of parking lots cannot keep pace with the multiplicity of cars. Parking problems are becoming increasingly serious. Currently in most parking lots, the management of entering and exiting automobiles consists of collecting fees at a pinpoint. This management style causes long queues at the inlet and outlet of the lot. Furthermore, parking management must usually satisfy certain requirements, such as passing speed and privacy protection of the various customers. Hence, automated management of parking lots is a pending problem needing resolution. Comparable to an identification card for each individual, a license plate is also unique for each automobile. If a license-plate recognition system could be implemented by applying the rapid processing capability of computers and a specific passing mechanism established for each type of parking customer, it would be possible to resolve several management problems regarding parking lots. In order to increase the efficiency of parking management, a mechanism is proposed to shorten the queue time for each automobile with different authority at the inlet and outlet of a parking lot by applying image analysis techniques and planning different ways to transit.