本文提出一個可應用於智慧型運輸系統之影像式停車資訊系統架構。系統利用基於LBP之影像式停車位使用狀況辨識方法，計算停車場影像之特徵，經過辨識與比對得到停車位使用狀況資訊。使用資訊轉換為易於傳輸之JSON格式，並予以簽章驗證，系統使用之驗證系統能抵抗傳輸錯誤、人為變造與重放攻擊，可利用網際網路傳輸至資訊整合主機，而不必要建置專用線路。資訊主機收取停車資訊後，依照停車區域的特性給予標籤，並利用標籤進行整合，增加整合的彈性並可依據目標地點提供加值停車資訊。所提出之使用狀況偵測法可應用於停車場管理、先進旅行者資訊系統以及交通管理使用。

Novel image-based parking space information architecture is proposed for intelligent transportation systems (ITS). An LBP-based recognition method is applied to detect parking space occupation status from images of parking lots. The detected occupation status information is converted into JSON format and signed. Signatures make the information robust enough to resist transmission errors, modifications and replay attacks. With signatures the information can be transmitted to a management server via internet instead of dedicated lines. By using internet and cameras originally in the parking lots, building costs of a parking space management system can be reduced. The management server integrates parking information by a smart labeling system. The labeling system increases the flexibility of the integrated information. The proposed system can be used in parking lot management, an ATIS, or a traffic management system.