Image Matching Scheme by using Bhattacharyya Coefficient Algorithm
A recurring problem that arises throughout the sciences is that of deciding whether two statistical distributionsdiffer or these are consistent - currently the chi-squared statistic is the most commonly used technique for addressing thisproblem. This paper explains the drawbacks of the chi-squared statistic for comparing measurements over largedistances in pattern space and suggests that the Bhattacharyya measure can avoid such difficulties. The originalinterpretation of the Bhattacharyya metric as a geometric similarity measure is reviewed and it is pointed out thatthis derivation is independent of the use of the Bhattacharyya measure as an upper bound on misclassification in a Two-class problem. The affinity between the Bhattacharyya measures is described and thatthe measure is applicable to any distribution of data. I explain that the Bhattacharyya measure is consistent withan assumption of a Poisson generation mechanism for individual measurements in a distribution which is applicableto a frequency (histogram) or probabilistic data set and suggest application of the Bhattacharyya measure to thefield of system identification.