在色彩調和理論的研究領域當中，以Moon and Spencer（1944
a~1944c）所提出的色彩量化模型最為著名，分別以色相差、明度差與彩
度差為量尺，對調和區與不調和區加以定義。但將人對色彩的視覺過程，
獨立成三個向度來考量，並不符合人的視覺經驗。因此，歐立成曾以均勻
色彩空間中兩色點的距離，定義同時涵蓋色相差、明度差與彩度差的綜合
色差（DE*ab），並以色光為顯色條件，探討色差與色面積因素對色彩調
和的影響，得到初步的結果。 有鑑於此，為了使這色彩調和理論的研
究更為完整，本研究將以物體色為顯色條件來繼續探討色彩調和，共包括
三項的研究主題：不同造形因素對色彩調和的影響、色彩差異與物體色色
彩調和的關係、以及色面積因素對物體色色彩調和的影響。 在第一項
研究主題中，我們以並置、穿插、包圍與不規則鄰接線這四種不同的造形
關係，當成測試圖形分別進行實驗，並分析其反應調和度之間的差異。實
驗結果顯示，四種造形關係所反應的色彩調和度，有顯著差異；而且以不
規則鄰接造形的影響程度最大，其它三種造形之間則無顯著差異。 第
二項研究主題，我們假設色彩差異與調和度之間存在某種函數關係。在實
驗中，以色票呈現400組不同色差的雙色配色圖形，讓受測者針對每一組
配色，進行調和度的評分。實驗結果顯示，色彩差異與色彩調和之間具有
三次曲線的函數關係；然而，其曲線特徵與色光色的實驗結果不同。
最後一項研究主題，我們假設在配色中，當組成色的面積與該組成色的強
烈程度成反比時，此配色才能調和。實驗同樣以色票來呈現色彩，記錄受
測者對於配色相同，但面積比不同的測試圖形，所反應的相對色彩調和度
。實驗結果顯示，色面積比的因素，對相對色彩調和度沒有明顯影響。這
與色光實驗的結果大致相同。 此外，針對傳統的配色原則作檢證的結
果顯示，當配色的明度對比很高時，其調和度較高。而當配色的色相相同
，但明度、彩度不同時（等色相配色），則調和度較高。若配色中，包含
觀者的喜好色，則該配色的調和度亦明顯提高。然而，其他的配色原則，
如「等值配色」等，在本研究中，並未有較高的調和度。

In the study of color harmony, Moon and Spencer have
proposed the most famous model, and defined the harmony/
disharmony interval with hue difference, value difference, and
chroma difference in a color combination. However, human does
not perceive a coloralong its three attributes independently..
For this reason, we have adopted the colordifference formula
(DE*ab) in the CIELAB color space as an integral color interval
toinclude hue, value and chroma difference of a color
combination in the previous study.We used this definition of
color interval to study the "influence of color interval and
area factor on color harmony among colors displayed on a CRT,
and got some elementary results. To expand the scope of the
previous study of color harmony, this study investigates color
harmony of object colors. There are three issues to be explored
in this study: 1.theinfluence of shape factor on color harmony;
2.the relationship between color interval and color harmony, and
3.the influence of area factor on color harmony. In the first
part of this study, four different shapes are adopted in the
experiment toreveal their influence on the perceived degree of
color harmony. The results show that the differences of the
degree of color harmony among these four shapes are
statistically significant. In the second part of this study,
we have tried to find out the degree of color harmony as the
function of DE*ab. In the experiment, 400 2-color combinations
with different values of DE*ab on color chips were presented to
the subjects to evaluate the degree of color harmony of each
combination. The results reveal a cubic-curve relationship
between color interval and color harmony, as the same in the
previous study, but the characteristic of this curve is somewhat
different with our previous result of color on CRT display.
Then, we have tested the hypothesis that a color combination
will be harmonious when the area ratio between the two component
colors is inverse to the "color intensity" ratio between them.
In the experiment, we recorded subjects'' responses of the
relative degree of color harmony to varied combinations with the
same component colors but different arearatios. The results show
that the factor of area ratio doesn''t obviously influence the
relative degree of color harmony. Finally, we have examined
the accordance between our data with some traditional principles
of color harmony, and summarized some guidelines for color
harmony.