Sajad Farokhi

Near infrared face recognition using Zernike moments and Hermite kernels

This work proposes a novel face recognition method based on Zernike moments (ZMs) and Hermite kernels (HKs) to cope with variations in facial expression, changes in head pose and scale, occlusions due to wearing eyeglasses and the effects of time lapse. Near infrared images are used to tackle the impact of illumination changes on face recognition, and a combination of global and local features is utilized in the decision fusion step. In the global part, ZMs are used as a feature extractor and in the local part, the images are partitioned into multiple patches and filtered patch-wise with HKs. Finally, principal component analysis followed by linear discriminant analysis is applied to data vectors to generate salient features and decision fusion is applied on the feature vectors to properly combine both global and local features. Experimental results on CASIA NIR and PolyU NIR face databases clearly show that the proposed method achieves significantly higher face recognition accuracy compared with existing methods.

Rotation and noise invariant near-infrared face recognition by means of Zernike moments and spectral regression discriminant analysis

Abstract. Face recognition is a rapidly growing research area, which is based heavily on the methods of machine learning, computer vision, and image processing. We propose a rotation and noise invariant near-infrared face-recognition system using an orthogonal invariant moment, namely, Zernike moments (ZMs) as a feature extractor in the near-infrared domain and spectral regression discriminant analysis (SRDA) as an efficient algorithm to decrease the computational complexity of the system, enhance the discrimination power of features, and solve the “small sample size” problem simultaneously. Experimental results based on the CASIA NIR database show the noise robustness and rotation invariance of the proposed approach. Further analysis shows that SRDA as a sophisticated technique, improves the accuracy and time complexity of the system compared with other data reduction methods such as linear discriminant analysis.

Recognition of Images Degraded by Gaussian Blur

In this paper, we propose a new theory of invariants to Gaussian blur. We introduce a notion of a primordial image as a canonical form of all Gaussian blur-equivalent images. The primordial image is defined in spectral domain by means of projection operators. We prove that the moments of the primordial image are invariant to Gaussian blur and we derive recursive formulas for their direct computation without actually constructing the primordial image itself. We show how to extend their invariance also to image rotation. The application of these invariants is in blur-invariant image comparison and recognition. In the experimental part, we perform an exhaustive comparison with two main competitors: 1) the Zhang distance and 2) the local phase quantization.

Near Infrared Face Recognition: A Comparison of Moment-Based Approaches

Moment based methods have evolved into a powerful tool for face recognition applications. In this paper, a comparative study on moments based feature extraction methods in terms of their capability to recognize facial images with different challenges is done to evaluate the performance of different type of moments. The moments include Geometric moments (GM’s), Zernike moments (ZM’s), Pseudo-Zernike moments (PZM’s) and Wavelet moments (WM’s). Experiments conducted on CASIA NIR database showed that Zernike moments outperformed other moment-based methods for facial images with different challenges such as facial expressions, head pose and noise.

Near infrared face recognition

As a primary modality in biometrics, human face recognition has been employed widely in the computer vision domain because of its performance in a wide range of applications such as surveillance systems and forensics. Recently, near infrared (NIR) imagery has been used in many face recognition systems because of the high robustness to illumination changes in the acquired images. Even though some surveys have been conducted in this infrared domain, they have focused on thermal infrared methods rather than NIR methods. Furthermore, none of the previous infrared surveys provided comprehensive and critical analyses of NIR methods. Therefore, this paper presents an up-to-date survey of the well-known NIR methods that are used to solve the problem of illumination. The paper includes a discussion of the benefits and drawbacks of various NIR methods. Finally, the most promising avenues for future research are highlighted.

Object tracking using SIFT and KLT tracker for UAV-based applications

In this paper, we propose a semi-automatic object tracking method based on a Scale Invariant Feature Transform (SIFT) and Kanade-Lucas-Tomasi (KLT) tracker. In our approach, the region of interest is specified by the user and then the interest points are detected. The tracker is then used to track the specified object in the consecutive frames. To overcome rapid changes of appearance, occlusion or disappearance from the camera view, we employ a forward-backward error compensation. Experimental results on VIVID dataset indicates that the proposed method has superior overall performance compared to more common methods in the field.

Accurate palm vein recognition based on wavelet scattering and spectral regression kernel discriminant analysis

Abstract. Palm vein recognition (PVR) is a promising new biometric that has been applied successfully as a method of access control by many organizations, which has even further potential in the field of forensics. The palm vein pattern has highly discriminative features that are difficult to forge because of its subcutaneous position in the palm. Despite considerable progress and a few practical issues, providing accurate palm vein readings has remained an unsolved issue in biometrics. We propose a robust and more accurate PVR method based on the combination of wavelet scattering (WS) with spectral regression kernel discriminant analysis (SRKDA). As the dimension of WS generated features is quite large, SRKDA is required to reduce the extracted features to enhance the discrimination. The results based on two public databases—PolyU Hyper Spectral Palmprint public database and PolyU Multi Spectral Palmprint—show the high performance of the proposed scheme in comparison with state-of-the-art methods. The proposed approach scored a 99.44% identification rate and a 99.90% verification rate [equal error rate (EER)=0.1%] for the hyperspectral database and a 99.97% identification rate and a 99.98% verification rate (EER=0.019%) for the multispectral database.