Real=time extraction of reference points from TV pictures - includes reference point generator with numerical analyser to extract spatial contours and movement detector for each point of image

摘要

The device comprises a numerical analyser (1) which identifies points in video images in terms of a matrix of rows and columns. The output from this is passed to three further components: a generator of spatial contours (3) which subejcts neighbouring points to comparisons of luminosity and colour (or shades of grey) according to predetermined thresholds, an image memory (2) and a movement detector (4) which detects variations in luminosity between successive images due to movement. Output from the contours generator (3) and the movement detector (4) are passed to a final component which generates reference points of the images (5) ready for transmission. USE/ADVANTAGE - Preparation of video images for transmission over telephone lines. Considerably reduces number of reference points required for transmission compared with other methods.