Performance of Histogram Modification by LOG Function for CBIR using Statistical Parameters of Bins Contents

Dr. H. B. Kekre; Kavita Sonawane

In this paper we are introducing the new histogram specification using LOG function to modify the histogram. This histogram modification brings positive change in the features to be extracted from R, G and B planes of the image modified using the LOG function. Feature vectors are of dimension 8 which are extracted from the image contents segregated into 8 bins. These 8 bins are obtained by partitioning each plane(R, G and B) histogram into two parts using ‘Centre of Gravity’ (CG). Image contents extracted into 8 bins are used to represent the feature vector in the form of statistical moments. First four moments namely Mean, standard deviation (STD), skewness (SKEW) and kurtosis (KURTO) are extracted separately for each color content of the pixels counted into 8 bins. Query and database Image comparisons are carried out using the three similarity measures Euclidean distance (ED), Cosine correlation distance (CD) and Absolute distance (AD). Experimentation is performed using database of 2000 BMP images having 20 different classes. Set of 200 query images is used to test and evaluate the performance of the proposed approach. Three performance evaluation parameters used to evaluate the system are Precision Recall Cross over Point (PRCP), Longest String (LS) and Length of String to Retrieve all Relevant (LSRR).

Select Volume / Issues:

• Volume 3 Issue 6, Nov. 2012

Year:

2012

Type of Publication:

Article

Keywords:

Histogram Specification; LOG; Bins; CG; Mean; STD; SKEW; KURTO; ED; CD; AD; PRCP; LS; LSRR

Journal:

IJECCE

Volume:

3

Number:

6

Pages:

1466-1471

Month:

November