葉片分類

Classification of Leaf images

李佳玲(LEE CHIA-LING)

指導教授 ： 陳淑媛

元智大學/資訊學院/資訊工程學系/碩士(2003年)

摘要

現有之影像查詢系統雖然不少，但大都是針對一般之影像資料庫進行查詢，較少有針對特定影像資料庫進行查詢，尤其是針對植物之資料庫更是少之又少。但植物在食用、藥用及工業上之用途早被肯定，近年來，它在生態環境中的地位更備受重視。但近年來許多植物被毀，全世界的種質資源正在下降。故讓更多的人們對植物有進一步的認識，進而珍惜愛護植物，是刻不容緩之課題。但要訓練民眾認識植物，除靠有限的專業植物學家外，更簡便的方式就是建立完整之植物資料庫，並提供方便有效之分類方式，以協助快速學習。
本論文以影像處理之技巧，建立完整之葉片資料庫，並提供方便有效之葉片分類法，以達到協助快速學習植物之目的。因現有文獻針對植物資料庫之查詢大都採用色彩及形狀特徵進行查詢，但因區域性形狀特徵一般而言較輪廓性形狀特徵穩定，故本論文提出一利用區域性形狀特徵進行葉片自動分類之新方法。本論文採用之特徵包括有葉子的長寬比例、葉子的邊緣點比例、葉子的中心點位置、及葉子水平垂直方向的投影。由眾多的實驗數據證實所提方法確實有效。

關鍵字

葉片資料庫 ； 影像分類 ； 投影 ； 葉片檢索

並列摘要

There are tremendous content-based retrieval systems. However, most of them are applied to general image databases. Only a few were proposed for specified databases such as satellite images, maps, faces, fingers and cultural relics. There are fewer for plant databases. The use of plant is plenty such as foodstuff, medicine and industry. Recently, plant is important for environment protection. However, the problem of plant destruction becomes worse in the few years. We should train people to know the plant, which in turn, to treasure and protect plant. In addition to the limited number of expert botanists, the convenient content-based retrieval system for plant is necessary and useful since it can facilitate fast learning of plants.
In this study, a leaf database is constructed and a classification method for the leaf database is proposed, which can facilitate fast learning of plant. Although most approaches used color and contour-based shape features, the proposed method tries to use region-based features. The reasons are that the region-based features are more reliable than the contour-based features. Those features include aspect ratio, edge ratio, region center position and the horizontal and vertical projections. The effectiveness of the proposed method has been demonstrated by various experiments.

並列關鍵字

leaf database ； content-based retrieval ； image classification ； projection ； region-based

參考文獻

[5] C. Im, H. Nishida, T.L Kunii, “Recognizing plant species by leaf shapes-a case study of the Acer family,” Pattern Recognition, Vol. 2, pp. 1171-1173, 1998.

[6] Z. Wang, Z. Chi, D. Feng, “Fuzzy integral for leaf image retrieval,” Fuzzy Systems, Vol. 1, pp.372-377, 2002.

[7] T. Saitoh, T. Kaneko, “Automatic recognition of wild flowers,” Pattern Recognition, Vol. 2, pp. 507 -510, 2000.

[8] D. Warren, “Automated leaf shape description for variety testing in chrysanthemums,” Image Processing and Its Applications, Vol. 2, pp. 497-501, 1997.

[9] S. Loncaric, “A survey of shape analysis techniques,” Pattern Recognition, vol. 32, no. 8, pp. 983-1001, 1998.