Open Access Open Access  Restricted Access Subscription or Fee Access

An Extensive Evaluation of Mean-Shift Tracking for Surveillance.

Darren Caulfield, Kenneth Dawson-Howe

Abstract


We evaluate the mean-shift tracker in the video surveillance domain in order to identify its best configuration. We compare the basic gradient ascent tracker with brute-force approaches. We assess the performance of multiple-part models and background exclusion (separately and combined). We assess its use with 21 colour spaces, 5 histogram quantisations, 6 convergence parameter values and 5 kernel types. We find that the best variant
of mean shift is gradient ascent with the Bhattacharyya coefficient, multiple-part models, background exclusion, HSV or YUV colour space, 4 bins per colour
channel, a convergence value of 0.25–1.00 and a cube root profile kernel. As a baseline, we include tracking using normalised cross correlation (NCC) and find that NCC appears to outperform all variants of the mean-shift tracker. In order to compare mean-shift and NCC tracking across colour spaces we also define a version of NCC for multi-channel data.

Keywords


Mean-shift tracking; normalised cross-correlation; multiple-part models; background exclusion; gradient ascent; Bhattacharyya coefficient techniques.

Full Text:

PDF


Disclaimer/Regarding indexing issue:

We have provided the online access of all issues and papers to the indexing agencies (as given on journal web site). It’s depend on indexing agencies when, how and what manner they can index or not. Hence, we like to inform that on the basis of earlier indexing, we can’t predict the today or future indexing policy of third party (i.e. indexing agencies) as they have right to discontinue any journal at any time without prior information to the journal. So, please neither sends any question nor expects any answer from us on the behalf of third party i.e. indexing agencies.Hence, we will not issue any certificate or letter for indexing issue. Our role is just to provide the online access to them. So we do properly this and one can visit indexing agencies website to get the authentic information. Also: DOI is paid service which provided by a third party. We never mentioned that we go for this for our any journal. However, journal have no objection if author go directly for this paid DOI service.