利用顏色字串與權重函數之京劇臉譜檢索

Color Strings and Weighting Function for Peking Opera Mask Retrieval

本研究針對京劇臉譜影像檢索做一相關研究。人類歷史的時間軸不斷地向前進，每一時期都有值得人們紀錄保存的事與物。我們的檢索方法可以對不同狀況下之京劇臉譜影像做很有效率的檢索。首先，我們將每個京劇臉譜的尺寸標準化為60X60像素，再用8條規則（利用RGB三原色之相對關係）將每個京劇臉譜轉換成顏色字串。其次，我們利用權重函數（權重可以作調整）進行比較要檢索的京劇臉譜的顏色字串與儲存於資料庫的顏色字串。最後，排名前10名相似的結果將被檢索出來。我們的系統特色為：1.基於內容的京劇臉譜檢索系統（基於相似性檢索）和2.基於文本的京劇臉譜檢索（非常迅速和成熟）。

In this paper, we demonstrate a hybrid system for efficient retrieval of Peking opera mask images under different conditions. First, we standardize each Peking opera mask to 60 x 60 pixels. Second, eight rule（using the relationships between the three primary colors of the RGB color model) are used to convert each Peking opera Mmask into a color string, and color strings stored in the database. Third, we compare the color strings of Peking opera mask images with color strings stored in the database using a weighting function. Finally, the ten most similar results are shown. Our system maintains the advantages of both content-based（based on similarity searching）, and text-based retrieval（rapid and mature）system.