Please use this identifier to cite or link to this item:
Title: A novel method for logo detection based on curvelet transform using GLCM features
Authors: Naganjaneyulu, G.V.S.S.K.R.
Sai, Krishna, C.
Narasimhadhan, A.V.
Issue Date: 2018
Citation: Advances in Intelligent Systems and Computing, 2018, Vol.704, , pp.1-12
Abstract: Automatic logo detection is a key tool for document retrieval, document recognition, document classification, and authentication. It helps in office automation as it enables the effective identification of source of a document. In this paper, a novel approach for logo detection using curvelet transform is proposed. The curvelet transform is employed for logo detection because of its ability to represent curved singularities efficiently when compared to wavelet and ridgelet transforms. The proposed algorithm consists of five steps, namely segmentation, noncandidate elimination, computation of curvelet coefficients, gray level co-occurrence matrix (GLCM) features extraction, followed by classification using a pretrained support vector machine classifier. The proposed algorithm is tested on a standard dataset, and the performance is compared with the state-of-the-art methods. The results show good improvement in the accuracy when compared with the competitors. � Springer Nature Singapore Pte Ltd. 2018.
Appears in Collections:2. Conference Papers

Files in This Item:
There are no files associated with this item.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.