Feature description based on Mean Local Mapped Pattern Carolina Toledo Ferraz * , Osmando Pereira Junior † , Adilson Gonzaga ‡ Department of Electrical Engineering - EESC/USP University of S˜ ao Paulo Av. Trabalhador S˜ ao Carlense, 400 13560-590, S˜ ao Carlos, SP, Brasil * caroltoledoferraz@gmail.com, † osmandoj@gmail.com, ‡ agonzaga@sc.usp.br Abstract—Local feature description has gained a lot of interest in many applications, such as texture recognition, image retrieval and face recognition. This paper presents a novel method for local feature description based on gray-level difference mapping, called Mean Local Mapped Pattern (M-LMP). The proposed descriptor is robust to image scaling, rotation, illumination and partial viewpoint changes. Furthermore, this descriptor more effectively captures the nuances of the image pixels. In our experiments, the descriptor is compared to the Center-Symmetric Local Mapped Pattern (CS-LMP) and the Center-Symmetric Local Binary Pattern (CS-LBP). The results show that our descriptor performs better compared to these two methods. I. I NTRODUCTION In image processing, the local feature description plays an important role in applications such as texture recognition, content-based image retrieval and face recognition. The objec- tive is to build a feature vector providing a representation that allows efficient matching of local structures between images. One of the most widely known key feature descriptors published in the literature is the scale-invariant feature trans- form (SIFT) [1]. It was introduced by Lowe in 1999 [2]. It is characterized by a 3D histogram of gradient locations and orientations and stores the bins in a vector of 128 positions. Many other descriptors like SIFT have been proposed, such as Principal Components Analysis SIFT (PCA-SIFT) [3], Gradi- ent Location and Orientation Histogram (GLOH) [4], Speeded Up Robust Features (SURF) [5], Colored SIFT (CSIFT) [6], Center-Symmetric Local Binary Pattern (CS-LBP) [7] and Kernel Projection Based SIFT (KPB-SIFT) [8]. The GLOH descriptor is very similar to SIFT, however, it uses a Log-Polar location grid instead of a Cartesian one. SURF approximates SIFT using the Haar wavelet response and using integral images to compute the histograms bins. It uses a descriptor vector of 64 positions, providing a better processing speed. Jin, Liu, Lu and Tong [9] proposed the Improved Local Binary Pattern (ILBP) that is an improvement of the LBP feature and is used for face detection [9]. It compares all of the pixels (including the central pixel) with the mean intensity. On the other hand, the Mean Local Binary Pattern (M-LBP) descriptor [10] does not consider the central pixel in this calculation. Center-Symmetric Local Binary Pattern (CS-LBP) [7] combines the strengths of the well-known SIFT descriptor and the Local Binary Pattern (LBP) texture operator. In [7], it is shown that the construction of the CS-LBP descriptor is simpler than SIFT, but it generates a feature vector of 256 positions, which is twice the SIFT vector size. In the Local Mapped Pattern (LMP) approach [11], the authors consider the sum of the differences of each gray- level of a given neighborhood to the central pixel as a local pattern that can be mapped to a histogram bin using a mapping function. The Center-Symmetric Local Mapped Pattern (CS-LMP) [11] combines the desirable properties of the CS-LBP and the Local Mapped Pattern (LMP). This approach is based on the sum of the differences of each gray-level of a given neighborhood to the center-symmetric pairs of pixels as a local pattern that can be mapped to a histogram bin using a mapping function. In this paper, we propose a new interest region descriptors, denoted as the Mean Local Mapped Pattern (M-LMP). The M-LMP descriptor is based on the CS-LBP, ILBP and M-LBP using the LMP methodology. This new descriptor captures the small transitions of pixels in the image more accurately, resulting in a greater number of correct matches than the CS- LMP and CS-LBP. The rest of this paper is organized as follows. In Section II, we briefly describe the CS-LMP, CS-LBP, ILBP, M-LBP and LMP. Section III gives details on the proposed approach. The experimental evaluation and the results are presented in Section IV. Finally, we conclude the paper in Section V. II. LOCAL DESCRIPTORS In this section, five previously published local descriptors are shown as the theoretic material for our approach: the CS- LBP, CS-LMP, ILBP, M-LBP and the LMP. A. Center-Symmetric Local Binary Pattern (CS-LBP) The CS-LBP is a modified version of the LBP texture feature and SIFT descriptor. In [7], the authors state that the LBP produces a rather long histogram and therefore is difficult to use in the context of an image descriptor. To solve this problem, they modified the scheme of how to compare the pixels in the neighborhood. Instead of comparing each pixel with the center one, the method compares center-symmetric pairs of pixels. For the construction of the descriptor, the image regions are divided into cells with a location grid, and for each cell, a DRAFT X Workshop de Vis˜ao Computacional - WVC 2014 32