Abstract :
A colour object search technique is presented. Given an image, regions of probabilistic occurrence are generated (for each object model) by isolating regions whose colours are similar to the template colours. Since the template may exist at one or more of these region locations, each is examined individually. At each region location the object size is estimated and an expansion process initiated to include all pixels matching the template colours. Expansion is terminated when a match measure quantified by defining a term `extent of match´ (EOM), based on object size and the number of pixels with each model colour, is maximised. If the EOM exceeds a predefined threshold then the object is assumed to be present. Several experiments are presented which demonstrate the algorithm´s robustness to scale, affine object distortion, varying illumination, image clutter and occlusion
Keywords :
image colour analysis; image matching; image segmentation; object recognition; search problems; affine object distortion; colour object search technique; expansion process; experiments; extent of match; illumination; image clutter; image occlusion; image regions; object model; object recognition; object size; pixels matching; region locations; regions of probabilistic occurrence; scale robustness; template colours; Fingerprint recognition; Histograms; Object recognition; Pixel; Robustness; Shape; Spatial resolution; Statistics; Tree graphs; Visual system;
