Design and Implementation of an Efficient Image Retrieval System using Spatial Feature Reduction
Iosr Journals, Naykude, Imran Dilavar, Prof. V. B. Gaikwad, Ravish R. Singh · Figshare · 2015
Image segmentation is a very emerging and important area due to a large number of real life applications. Bag-of-features (BoFs) model is the one of the most successful algorithm used for the image classification. BoF methods are based on order-less collections of quantized local image descriptors; they discard spatial information and are therefore conceptually and computationally simpler than many alternative methods.Bag-of-features algorithm is used for the feature extraction from the image. BoFs model has several advantages like,scalability,simplicity, and generality. But BoFs have some disadvantages too and some of them are Time and Accuracy of the image classification process. In this paper Bag-of-features model is extended by using spatial pooling to improve the time and accuracy of the image classification model. In proposed method first the system is trained by creating the database of the image feature for the evaluation process and then the evolution of the features is done for the input image by using clustering algorithm. In the proposed method KKN algorithm is used to find out the similar features from the database. Spatial pooling is then used to improve the performance of the existing system.