This paper proposes Object Based Image Retrieval (OBIR) System with segmenting the objects from the images and then extracting various features from the objects. The objects are the most prominent part of an image which relates more to the human perception. First, the object present in the images is segmented by four different segmentation techniques such as K-means, Active Contours, Edge-Convex hull and Global Thresholding. Later, the color features such as Color Histogram (CH) and Color Coherence Vector (CCV), Texture feature using Local Binary Patterns (LBP) and shape feature using Histogram of Gradients (HOG) are extracted. Finally, with the usage of different segmentation and techniques mentioned above feature are extracted from objects. Results obtained are tabulated and performance study is made.
A Neuromorphic Proto-Object Based Dynamic Visual Saliency Model with a Hybrid FPGA Implementation