應用不受光照影響之色彩重心於影像索引

Image Indexing with Illumination Invariant Color Centroids

張倉瑜(Tsang-Yu Chang)

指導教授 ： 黃惠藩

亞洲大學/資訊學院/資訊工程學系碩士班/碩士(2011年)

摘要

龐大且多種類的圖片資料庫內搜尋圖片是不易的，因此有許多影像索引的方法相繼被提出。影像成像時又易於受到光照的影響而改變，因此克服光照的影響是重要的議題。許多先進們發表了克服光照問題的方法都建立於三維的色彩空間模型上。本論文提出一個不受光照影響的色彩重心應用於影像索引。本論文的方法將三維的色彩空間轉換到二維的色彩重心代表，此色彩重心代表提供了壓縮的色彩資訊且對於光照的變動比較不敏感。基於色彩不變模型，我們所提出的方法不會受到光照條件改變的影響。實驗的結果顯示在影像索引上用新的色彩重心模型較不易受到光照的改變影響。

關鍵字

影像索引 ； 色彩重心代表 ； 不受光照影響

並列摘要

This paper presents an illumination invariant color centroid representation for image indexing. The proposed method transforms 3-D color space to a 2-D color centroid representation which provides compact color information representation and is less sensitive to illumination variation. Based on color constancy model, the proposed color centroid representation is invariant with respect to illumination variation. Experimental results show robust image indexing using the new color centroid model on objects that are illuminated under different illumination conditions.

並列關鍵字

Image Indexing ； Color Centroid Representation ； Illumination Invariant

參考文獻

[1]　H. Ng, "Illumination and highlights invariant color model for image matching," The Imaging Science Journal, vol. 57, no. 3, pp. 167-173, 2009.

[2]　T. Gevers and H. Stokman, "Robust histogram construction from color invariants for object recognition," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 26, no. 1, pp. 113-118, 2004.

[4]　H. Smithson and Q. Zaidi, "Colour constancy in context: Roles for local adaptation and levels of reference," Journal of Vision, vol. 4, no. 9, pp. 693-710, 2004.

[5]　Q. Zhang, S. Kamata, and J. Zhang, "Color barycenter hexagon model based road sign detection," in 2008 International Multi-Conference of Engineers and Computer Scientists (IMECS 2008), vol. 1, Hong Kong, 2008, pp. 655-658.

[7]　Y. Gong, H. Chua, and X. Guo, "Image indexing and retrieval based on colour histograms," International Journal of Multimedia Tools and Applications, vol. 2, no. 2, pp. 133-156, 1996.