Coalescing color texture and shape texture for an efficient Image Retrieval

It is crucial to create indexing systems, in order to properly access the images, due to the volume of images in the digital world. Content Based Image Retrieval (CBIR) facilitates the indexing to retrieve similar images. This article suggests a CBIR that combines basic attributes from the image for instance shape, color, and texture to retrieve the alike images. The color texture attribute is extracted from the image based on thresholding over color plane and local binary pattern. The shape texture feature is represented by legendre moments from local binary pattern images obtained through wavelet planes. The combination of color texture and shape texture information makes the efficient image representation, which in turn effectively retrieves the similar images from the large set. On the Corel-1K dataset, the suggested approach has been evaluated. Recall and precision measures have been applied to assess the proposed method's effectiveness. According to the experimental findings, the coalesced method based image retrieval system gives notable improvements as compared to systems use either one of color texture or shape texture features.