Retrieval Architecture with Classified Query for Content Based Image Recognition

Table 2

Paired t-test for significance testing of precision results for classification.

Comparison	-calc	value	Significance of difference in value
Feature extraction by binarization using bit plane slicing with Niblack’s local threshold method (Thepade et al., 2014 [9])	3.1626	0.0133	Significant
Feature extraction by binarization with multilevel mean threshold (Kekre et al., 2013 [19])	3.1626	0.0133	Significant
Feature extraction by binarization using Bit Plane Slicing with mean threshold (Kekre et al., 2013 [19])	2.9059	0.0197	Significant
Feature extraction by binarization of original + even image with mean threshold (Thepade et al., 2013 [20])	3.4116	0.0092	Significant
Traditional feature extraction by binarization with Bernsen’s local threshold method (Yanli and Zhenxing, 2012 [26])	3.2593	0.0115	Significant
Traditional feature extraction by binarization with Sauvola’s local threshold method (Ramírez-Ortegón and Rojas, 2010 [23])	3.0157	0.0167	Significant
Traditional feature extraction by binarization with Niblack’s local threshold method (Liu, 2013 [22])	3.9038	0.0045	Significant
Traditional feature extraction by binarization with Otsu’s global threshold method (Shaikh et al., 2013 [24])	3.661	0.0064	Significant