隨著電腦科技的突飛猛進,以及數位影像處理技術的突破與廣泛應用,美容整體造型設計的發展不得不跳脫傳統的陰影而走向科技時代的潮流。本研究目的在於應用影像處理技術,發展一套臉型判別系統模式,以利於美容整體造型設計的發展。此系統模式利用臨界值法、影像外形處理與影像投影等基本影像處理自動找出特徵點的座標後,計算相對應的特徵向量組合,並依臉型長與寬之比例關係找出其對應的權值。在比對方面,依系統所找出的特徵值將可以有效代表臉部特徵,接著利用歐基里德距離及漢明距離來完成比對工作,也就可以完成臉型判別的結果。本研究最後以實際人臉影像進行實驗,其結果顯示以影像處理技術於臉型判別上有很高的辨識率。
The great advances in computer technology and digital image processing techniques have helped cosmetology styling design and education to escape from the shadow of tradition. In order to advance the development of cosmetology styling designing, the purpose of this research is by the application of image processing techniques to develop a system to assist in the decision on how a face should be shaped. In this system, the basic image processing techniques which will include the thresholding method, image forming treatment, image projection etc., are first used to locate the coordinates of feature points automatically. Then their corresponding eigenvectors are calculated as features, and the corresponding weights can be found by using the ratio of length of the face to its breadth. As regards comparison between features, both the Euclidean and Hamming distances are used to compare facial features, and then the decision on the shape can be made. Finally, this research carries out experiments with real facial images and the results show that there are high distinguishing rates in the system proposed in this research into decision-making in regard to face-shaping.