B-spline over-complete wavelet based fractal signature analysis for texture image retrieval

In the paper, we proposed a novel over-complete B-Spline wavelet based statistical features and fractal signatures for texture image analysis and retrieval. The discrete wavelet frame took the first order derivative of smoothing function into account, which is equivalent to Canny edge detection, with the specific case using Gaussian function as smoothing function. Meanwhile, the feature set based on the fractal surface area function in a Besov space is very accurate and robust for gray scale texture classification. Experimental results have shown that the proposed method is reasonable to describe the characteristics of the texture in temporal-frequent and fractal domain and it can reach the highest retrieval rate comparing with Gabor Filter based feature descriptor and B-Spline over-complete wavelet transformation based feature representation only.