1 Hybrid Local and Global descriptor enhanced with Color Information Leila Kabbai 1 , Mehrez Abdellaoui 1 , Ali Douik 2 1 National Engineering School of Monastir- ENIM, University of Monastir-Tunisia 2 National Engineering School of Sousse- ENISo, University of Sousse-Tunisia kabbai.leila@gmail.com, mehrez.abdellaoui@enim.rnu.tn, ali.douik@enim.rnu.tn Abstract: Feature extraction is one of the most important steps in computer vision tasks like object recognition, image retrieval and image classification. It describes an image by a set of descriptors where the best one gives a high quality description and a low computation. In this paper, we propose a novel descriptor called Histogram of Local and Global features using Speeded Up Robust Feature (SURF) descriptor (HLG SURF ) based on a combination of local features obtained by computation of Bag of Words of SURF and global features issued from a novel operator called Upper and Lower Local Binary Pattern (UL LBP ) that encodes the texture analysis associated with Wavelet Transform. To enhance the effectiveness of the descriptor, we used the color information. To evaluate the proposed method, we carried out experiments in different applications like image retrieval and image classification. The performance of the suggested descriptor was evaluated by calculating both Precision and Recall values using the challenging Corel and COIL-100 datasets for image retrieval. For image classification, the performance was measured by the classification rate using the challenging Corel and MIT scene datasets. The experimental results showed that the proposed descriptor outperforms the existing state of the art results. Key words: SURF, SIFT, Wavelet Transform, BoW, LBP. 1. Introduction Visual content description is one of the most challenging tasks in computer vision, like image retrieval [1], image classification [2], object recognition [3] and image matching [4]. Over the last few years much effort has been made in developing new features to yield a good classification [5]. These features can be local or global. The global features describe the visual content of the whole image. The global descriptor represents an image by a single vector. It is sensitive to lighting changes and view point and fails to identify the most important features of the image. Global features are unsuitable for some applications. Their drawbacks are resolved by the local features which are based on Interest Points (IPs). These IPs extract the best local image information [6]. Different approaches have been proposed to detect IPs such as Harris and Stephens [7], Harris-Lapace [8] and Scale Invariant Feature Transform (SIFT) [3]. Mikolajczyk and Schmid [8] reported an experimental evaluation of several descriptors and found that the SIFT algorithm obtained the best matching results when applying geometric and photometric transformations. Various extensions of SIFT descriptor were developed, for Page 1 of 21 IET Review Copy Only IET Image Processing This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication in an issue of the journal. To cite the paper please use the doi provided on the Digital Library page.