Interactions between cues to motion in depth

物体朝向观察者的运动由以下信息来源(线索)表示：(1)双眼图像的绝对视差的变化以及由此产生的眼睛的会聚运动。(2)物体远近区域的相对双眼视差(内部视差)的增加。(3)改变物体和静止物体之间的相对双目视差。(4)物体形象的隐现。(5)物体远近区域之间的运动视差。(6)住宿条件的变化。我设计并制造了一种独特的装置，使我能够独立控制上面的每一条线索。两个相同的图像对象(纹理表面或真实的3-D对象)被放置在镜面立体镜的两个臂中，并在手臂够得着的范围内沿着线性轨道来回移动。图像对象融合以创建朝向和远离观察者的立体对象。通过旋转每个轨迹来控制绝对视差(收敛)。通过围绕每个图像对象的中垂直轴旋转每个图像对象来控制立体对象的内部视差。通过使图像对象保持一段距离来控制隐现，同时通过横向移动图像对象来改变绝对视差，或者通过旋转图像对象来改变相对视差。因此，我可以使用真正深度移动的真实对象来更改任何深度运动线索或任何线索组合。人们普遍认为，绝对视差的变化不会产生深度运动的感觉。我已经用新的设备证明了这种假设是不正确的。关键因素是是否存在与视差信号相矛盾的图像隐约可见。固定纹理正面表面上的内部视差的处理已知会影响表面的感知距离和形状。我将使用新的仪器来研究真实表面和三维物体深度移动时内部视差的距离缩放。仅当假设对象为刚性时，内部视差的变化才是向纵深运动的提示。我将通过使用熟悉的物体、抽象的物体和清楚地表达的物体来研究假定的僵化的作用。

The motion of an object towards an observer is indicated by the following sources of information (cues): (1) Changes in the absolute disparity of the images in the two eyes and the vergence movements of the eyes that these create. (2) An increase in the relative binocular disparity of the near and far regions of the object (internal disparity). (3) Changing relative binocular disparity between the object and stationary objects. (4) Looming of the image of the object. (5) Motion parallax between the far and near regions of the object. (6) Changes in accommodation. I have designed and built a unique apparatus that allows me to independently control each of the above cues. Two identical image objects (textured surfaces or real 3-D objects) are placed in the two arms of a mirror stereoscope and moved to-and-fro along linear tracts within arm's reach. The image objects fuse to create a stereo object moving towards and away from the observer. Absolute disparity (vergence) is controlled by rotating each track. The internal disparity of the stereo object is controlled by rotating each image object about its mid-vertical axis. Looming is controlled by keeping the image objects at one distance while changing absolute disparity by moving the image objects laterally or relative disparity by rotating them. I can thus change any motion-in-depth cue or any combinations of cues using real objects really moving in depth. It is generally believed that changes in absolute disparity do not create sensations of motion in depth. I have already shown with the new apparatus that this assumption is incorrect. The crucial factor is whether or not there is image looming that contradicts disparity signals. Manipulations of internal disparity over a stationary textured frontal surface are known to affect the perceived distance and shape of the surface. I will use the new apparatus to investigate the distance scaling of internal disparity with real surfaces and 3-D objects moving in depth. Changes in internal disparity are a cue to motion in depth only when the object is assumed to be rigid. I will investigate the role of assumed rigidity by using familiar objects, abstract objects and objects that articulate.