| 摘要 |
<p>The purpose of the present invention is to match zoom features of left and right cameras, if zoom features of each camera do not match in the process of photographing a stereoscopic image by using a rig mounted with left and right cameras. In the case of photographing a stereoscopic image, basically, left and right cameras should be prepared and an image should be photographed simultaneously. Left and right cameras are independent cameras and are controlled by means of motor controlling provided inside in respective cameras and therefore, there is a margin of error, physically. Unless left and right cameras have perfectly same zoom features, upon synthesized with a stereoscopic image, the size of an object does not match due to differences of each zoom, and as a consequence, a visual dizziness occurs. The purpose of the present invention is to derive a relative size of an image by using feature points of an incoming image from left and right cameras and if necessary, to minimize deviations occurring between cameras by means of correction. That is, the features of a same object are detected in respective images photographed by left and right cameras, relative distances and sizes existing between the same object are analyzed, and thereby, the deviations of zoom features in left and right cameras are discovered, the deviations are reflected automatically or manually, and thereby, zoom features of left and right cameras can be matched and accordingly, a visual fatigue related to a stereoscopic image can be minimized.</p> |
