| 摘要 |
<p>The invention relates to an image processing method, for constructing a so-called target image (Io), from adjacent images, with fixed framing, called source images (I1, ... Ij, ... In), having substantially coincident vanishing points. This method comprises the steps of: digitising of the images; determination, for one of the pixels of the target image (Io) of the address (Aq) of a so-called corresponding point in a source image (Ij) among all the source images; determination of the value (F) of the luminance at this corresponding point; attribution to the starting pixel in the target image (Io) of the luminance value (F) of this corresponding pixel; and repetition of these steps for each of the pixels of the target image (Io). The invention also relates to a device implementing this method and comprising a system of n real fixed cameras (C1, ... Cn) which supply n adjacent source images (I1, ... In) covering a wide-angle field of view, and which have common vanishing points (P); and an image construction system (100) simulating a mobile, so-called virtual camera (Co), for supplying a so-called target sub-image (Io), of the wide-angle field of view, and constructed from source images with the same vanishing point (P). Application: surveillance devices. <IMAGE></p> |
