Fusion of Visible and Infrared Images using Empirical Mode Decomposition to Improve Face Recognition

Abstract

In this effort, we propose a new image fusion technique, utilizing Empirical Mode Decomposition (EMD), for improved face recognition. EMD is a non-parametric datadriven analysis tool that decomposes non-linear nonstationary signals into Intrinsic Mode Functions (IMFs). In this method, we decompose images from different imaging modalities into their IMFs. Fusion is performed at the decomposition level and the fused IMFs are reconstructed to form the fused image. The effect of fusion on face recognition is measured by obtaining the Cumulative Match Characteristics (CMCs) between galleries and probes. Apart from conducting face recognition tests on visible and infrared raw datasets, we use datasets fused by averaging, principal component (PCA) fusion, wavelet based fusion and our method, for comparison. The face recognition rate due to EMD fused images is higher than the face recognition rates due to raw visible, raw infrared and other fused images. Examples of the fused images and illustrative CMC comparison charts are shown.

Facial emotion recognition using empirical mode decomposition

“…Recent studies show that EMD algorithm was used to detect epileptic seizure using EEG signals (Kaleem, Guergachi, & Krishnan, 2013;Orosco, Lacia, Correa, Torres, & Graffigna, 2009;Pachori, 2008) and diagnose Alzheimer disease (Gallix, Gorriz, Ramirez, & Lang, 2012). Also, EMD algorithm is extended into bi-dimensional EMD (BEMD) and has been used for texture analysis (Nunes, Bouaoune, Delechelle, Niang, & Bunel, 2003), image compression (Linderhed, 2005), skeletonization pruning (Krinidis & Krinidis, 2013), image fusion (Hariharan, Koschan, Abidi, Gribok, & Abidi, 2006), face recognition (Bhagavatula & Savvides, 2007;Zhang & Tang, 2009) and facial pose-estimation (Qing, Jiang, & Yang, 2010). EMD is a multi-resolution decomposition technique suitable for nonlinear data has been used to decompose any complicated signal into frequency components called IMFs.…”

Investigating 3D echocardiography image fusion for improving image quality

“…Diseases related to heart and blood circulatory system are termed as cardiovascular disease (CVDs) [10]. In 2009, according to an estimate, over 180,000 people died from CVDs, which form one third of total deaths in the UK.…”

Extraction of Illumination-Invariant Features in Face Recognition by Empirical Mode Decomposition

“…One is to view the image as a long lengthened vector and then apply EMD on it. Another one is to apply on each row or column one by one as literatures [4,5,6,7] did. In order to discriminate these two methods, we name the former one as 1DEMDa and the latter one as 1DEMDb.…”

“…The image data has been expressed in terms of an array of rows and columns, then the EMD is applied to these arrays row by row. Linderhed et al also adopted this row-by-row EMD method in the fusion of visible and infrared images [5,6]. The input images are vectorized in lexicographical order and EMD is performed on each channel vector separately.…”