Title :
A method to reconstruct occluded contour of a partially occluded circle
Author_Institution :
Queensland Univ. of Technol., Brisbane, Qld., Australia
Abstract :
A object representation-magnitude of edge pixel pointing vector (MEPPV) which is up to similarity invariant is developed. An array of MEPPV can be utilized to recognize partially occluded object with unknown position, orientation and scale, and to retrieve the occluded contour of this object. The matching of the occluded object with the model is a global matching. For a circular object, the influence of the projective transformation on the MEPPV is analysed through estimating the ratio of major axis to minor axis of the ellipse in the image as well as the scale. Experimental results are presented
Keywords :
edge detection; image matching; image reconstruction; image representation; object detection; circular object; ellipse; experimental results; global matching; magnitude of edge pixel pointing vector; major axis to minor axis ratio; object representation; occluded contour reconstruction; occluded object matching; orientation; partially occluded circle; position; projective transformation; scale; similarity invariant representation; Australia; Fasteners; Image recognition; Image reconstruction; Machine vision; OFDM modulation; Pixel; Robotic assembly; Shafts; Shape;
Conference_Titel :
Signal Processing, 1996., 3rd International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-2912-0
DOI :
10.1109/ICSIGP.1996.566287
