Hybrid feature to encode shape and texture for Content Based Image Retrieval

Abstract

In this paper, we propose an approach for representing both shape and texture information in an image using a single hybrid feature descriptor for Content Based Image Retrieval. Towards this, we compute the gradient magnitude of the input image prior to deriving features. Feature extraction is then performed using the responses from a bank of Gabor filters. Here, we exploit the fact that shape corresponds to the high spatial frequency content in the image whereas natural texture information predominantly lies within low to mid-range frequencies. This approach helps in better localization of characteristic texture as well as shape, due to spread of energy towards high frequencies in spectral domain. Moment invariants are extracted from Gabor filter responses which yield better retrieval performance than conventional statistical features. Experimental results show that this approach has relatively improved retrieval performance on Corel image data set when compared with recent approaches in the literature. Further experiments were also performed on a medical image dataset with 95.4 percent precision and 74.6 percent recall.

PLT‐based spectral features for texture image retrieval

“…This affects the CBIR performance. Curvelet is more effective in capturing curvilinear properties such as lines and edges [41,42]. Curvelet obeys parabolic scaling.…”

Multimodel Local Terenary Patterns for Content Based Image Retrieval

“…Rai, H.G.N. et al[10] "Hybrid feature to encode shape and texture for Content Based Image Retrieval" In this paper, author was proposed a methodology for speaking to both shape and composition data in a picture utilizing a solitary half breed gimmick descriptor for Content Based Image Retrieval.…”