頁籤選單縮合
題名 | 相對像差與整體一致性在立體攫取中所扮演的角色 |
---|---|
作者 | 李宏鎰; | 書刊名 | 應用心理學報 |
卷期 | 3 1994.09[民83.09] |
頁次 | 頁67-119 |
分類號 | 176.12 |
關鍵詞 | 相對像差; 立體攫取; |
語文 | 中文(Chinese) |
中文摘要 | 本研究主要的目的在於提出像差相對性的觀念,來闡述立體攫取之所以會發生的 原因為何。根據「圖形具有正像差者上浮;負像差者下沉;零像差者不浮不沉」的原則。本 研究做了一系列的實驗,以澄清 Ramachandran 於 1986 年所提出的立體攫取現象,乃是因 錯覺輪廓內的壁紙圖案, 相對於誘發元素外緣具正像差, 所以壁紙圖案會上浮, 而非如 Ramachandran 所言,壁紙圖案為零像差。 本研究的思路是先闡明像差是必須以相對性的角 度來看待,以此抽絲剝繭地說明立體攫取發生的原因,乃在於有相對像差線索存在所致。接 下來,從真實輪廓與錯覺輪廓所製成的刺激材料之比較中,了解到 Ramachandran 所使用之 刺激材料,恰好可符合 Julesz ( 1960 )所提出的立體圖對製作原理,即左右眼圖各有一 不能配對的單眼區;由此點也可以看出,立體攫取之所以會發生乃是由於被攫取的壁紙圖案 本身已具有像差所致。為了更仔細地了解 Ramachandran 所用之刺激材料的特性,本研究尚 操弄以隨機點所構成的壁紙圖案之刺激材料。以此說明,倘若背景無法被輪廓分割為兩個獨 立部份,則立體攫取將不可能發生,即使是在真實輪廓的情況下亦然。由此亦可看出壁紙圖 案本身週期性的特色有助於立體攫取的發生。實驗一是以像差相對性的概念來說明立體攫取 發生的原因;實驗二則在比較隨機點與壁紙圖案特性之差異;實驗三是分辨真實輪廓與錯覺 輪廓於立體攫取中的角色異同; 實驗四則在比對 Mather ( 1989 )與 Ramachandran ( 1986 )之實驗結果;實驗五是對於立體攫取之相位特性做進一步的探討。 總結本研究所得 到的結果;首先,立體攫取之所以會發生,仍是因為輪廓內的壁紙圖案相對於誘發元素外緣 具有像差所致。其次,隨機點不具有壁紙圖案的周期性特性,不能被分割成兩獨立部份,故 以隨機點為背景,不會有立體攫取現象。再者,相位會影響立體攫取的有無,但立體攫取若 要發生非得要有相對像差的存在不可。本研究所要強調的是:「像差」在雙眼配對所形成的 立體知覺中,仍然扮演著決定性的角色。 |
英文摘要 | PURPOSE. Ramachandran (1986) showed that stereo capture occurs when inducing elements are superimposed on a wallpaper pattern of vertical lines. The pattern within the subjective figure (inner pattern) is captured with the contour and perceived as being in front of the rest of the pattern. Because inner pattern is at the same depth plane as the rest of the plane, he argued that capture occurs in the context of "zero disparity". The purpose of the present studies is to demonstrate that disparity cues exist in Ramachandran's stimuli and that disparity cues alone can produce stereo capture. MATHODS & RESULTS. Disparity relativity is defined as the relative horizontal disparity between the two eye-views in the distance from the first line of the inner pattern to one of the two cues:(1) the outer boundary of the inducing elements, and (2) the outer boundary of the background. The hypothesis is that the first should be sufficient to produce stereo capture. To verify this hypothesis, we first used wallpaper patterns of vertical lines without the outer boundary of the background. All of the 12 subjects reported stereo capture. We then set the disparity relativity to be zero from the first line of the inner pattern to the outer boundary of the inducing elements. Under such conditions when neither of the disparity cues exists, stereo capture did not occur. Similar results were found with patterns of horizontal lines or with real contours. We then shifted the whole inner pattern of one eye-view while keeping the other eye-view constant. At zero shift, where no disparity existed in the distance from the first line of the inner pattern to the outer boundary of the inducing elements, stereo capture did not occurs. As phase shifted a cycle and the previous second line of the shifted pattern became the reference for matching with theother eye-view, stereo capture emerged. Stereo capture was reported when phase was the same and disparity existed in the two eye-views from the first line of the inner pattern to the outer boundary of the inducing pattern. When phase was different, stereo capture was unstable. CONCLUSIONS. relative disparity is necessary to produce stereo capture. Phase of the inner pattern is also important. When relative disparity exists between the two eye-views from the first line of the inner pattern to the outer boundary of the inducing patterns or to the outer boundary of the background pattern, stereo capture occurs. When such relative disparity does not exist in wallpaper patterns or in patterns of random dots, capture does not occur. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。