Content Based Image Retrieval using Color-Edge Detection and Haar Wavelet Transformation
Published: 2016
Author(s) Name: Dileshwar Patel, Amit Yerpude |
Author(s) Affiliation: Computer Science & Engineering Rungta College of Engg & Technology Bhilai, Chhattisgarh, India
Locked
Subscribed
Available for All
Abstract
Content based image retrieval is one of the emerging field that continuously broader their research area. CBIR basically uses color as the fundamental feature to get the internal aspect of an image. However image retrieval using only color feature is not sufficient to retrieve all the relevant images. Therefore we need to combine two or more features to make the performance of the system better. CBIR includes other low level features such as shape and texture based. As the solution of the above problem this paper describes a novel algorithm that provides the straight forward method to get the higher retrieval performance. In this method color edge detection and haar wavelet transformation is used. The proposed algorithm is different from previous method since this method combines the both color and edge features. Also the feature size is reduced or normalized with the help of haar wavelet transform. Performance is tested by altering the query images. For Image set we are using wang image database. Finally the experimental results are evaluated based on precision and recall value.
Keywords: CBIR, Feature Extraction, Colour, Shape, Texture, Edge, Haar Wavelet, Canny
View PDF