Image Class Ranking using Bag of Features
Sreedevi Kulkarni, T.M. Manu · 2022 2nd International Conference on Intelligent Technologies (CONIT) · 2022
Selecting a suitable approach for image ranking or classification will form the basis of any computer vision application. Bag of features also called Bag of visual words or just visual words approach is the most used method for image classification even today. Bag of visual words is used to predict the category of images in the dataset. Feature vectors are the histogram of visual occurrences for this approach. The benchmark dataset used for validation is Caltech-101 and Flower dataset. Experimentation shows three different ways of partitioning the training and validation data. Depending on the deduced results, the performance of category wise results is calculated.