A New Hybrid Descriptor Based on Spatiogram and Region Covariance Descriptor

摘要

Image descriptors play an important role in any computer vision system e.g. objectrecognition and tracking. Effective representation of an image is challenging due tosignificant appearance changes, viewpoint shifts, lighting variations and varied objectposes. These challenges have led to the development of several features and theirrepresentations. Spatiogram and region covariance are two excellent image descriptorswhich are widely used in the field of computer vision. Spatiogram is a generalization of thehistogram and contains some moments upon the coordinates of the pixels corresponding toeach bin. Spatiogram captures richer appearance information as it computes not onlyinformation about the range of the function like histograms, also information about the(spatial) domain.However, there is a drawback that multi modal spatial patterns cannot bewell modelled.Region covariance descriptor provides a compact and natural way of fusingdifferent visual features inside a region of interest. However, it is based on a globaldistribution of pixel features inside a region and loses the local structure.In this paper, weaim toovercome the existing drawbacks of these descriptors. To this, we propose rspatiogramand then a new hybrid descriptor is presented which is combination of rspatiogramand traditional region covariance descriptors. The results show that ourdescriptors have the discriminative capability improved in comparison with otherdescriptors.