Colour logo and trademark retrieval and tracking in unconstrained image sequences using Colour Edge Gradient Co-occurrence Histograms

摘要

In this paper, we present an extension of the colour edge co-occurrence histogram (CECH) object detection scheme for retrieving and tracking colour logos and trademarks in unconstrained image sequences. We introduce more accurate information to the CECH by virtue of incorporating colour edge detection using vector order statistics, producing more accurate representations of edges in images, as compared to the simple colour difference edge classification performed in the CECH. Our proposed method is thus reliant on edge gradient information, thus naming it the colour edge gradient co-occurrence histogram (CEGCH). We also use a colour quantization scheme based in the hue-saturation-value (HSV) colour space, for it is more suitable for logo and trademark retrieval, in comparison to the colour quantization scheme used with the CECH. Results illustrate that the CEGCH retrieves and tracks logos and trademarks with greater accuracy in comparison to the CECH.