Design of a Feature Descriptor for Content Based Image Retrieval. Image Retrieval

      Аннотация: Content based image retrieval is a technique for retrieving similar images for a particular query image from a large image dataset. Different types of image have different features. Extracting image features according to the user requirement from the dataset is a difficult task in the present scenario. Many feature descriptors have been proposed for image retrieval in the past few years. Image retrieval based on the content relies heavily not only the type of descriptors, but on the steps taken further. Most of them stick to a particular type of images. So, a multipurpose image feature descriptor is the essential need of the present scenario which can be beneficial to the different kind of images.

Ключевые слова: CBIR, MRELBP, Dominant Rotated Local Binary Pattern(DRLBP), Paricle Swarm Optimization(PSO), Support Vector Machine(SVM), classification, Deep Neural Networks(DNN), Neuro-Fuzzy Classifier(N-F), Machine Learning(ML), image retrieval, Feature Extraction.