We introduce a new model for personal recognition based on the 3-D geometry of the face. The model is designed for application scenarios where the acquisition conditions constrain the facial position. The 3-D structure of a facial surface is compactly represented by sets of contours (facial contours) extracted around automatically pinpointed nose tip and inner eye corners. The metric used to decide whether a point on the face belongs to a facial contour is its geodesic distance from a given landmark. Iso-geodesic contours are inherently robust to head pose variations, including in-depth rotations of the face. Since these contours are extracted from rigid parts of the face, the resulting recognition algorithms are insensitive to changes in facial expressions. The facial contours are encoded using innovative pose invariant features, including Procrustean distances defined on pose-invariant curves. The extracted features are combined in a hierarchical manner to create three parallel face recognizers. Inspired by the effectiveness of region ensembles approaches, the three recognizers constructed around the nose tip and inner corners of the eyes are fused both at the feature-level and the match score-level to create a unified face recognition algorithm with boosted performance. The performances of the proposed algorithms are evaluated and compared with other algorithms from the literature on a large public database appropriate for the assumed constrained application scenario.

We introduce a novel multimodal framework for face recognition based on local attributes calculated from range and portrait image pairs. Gabor coefficients are computed at automatically detected landmark locations and combined with powerful anthropometric features defined in the form of geodesic and Euclidean distances between pairs of fiducial points. We make the pragmatic assumption that the 2-D and 3-D data is acquired passively (e.g., via stereo ranging) with perfect registration between the portrait data and the range data. Statistical learning approaches are evaluated independently to reduce the dimensionality of the 2-D and 3-D Gabor coefficients and the anthropometric distances. Three parallel face recognizers that result from applying the best performing statistical learning schemes are fused at the match score-level to construct a unified multimodal (2-D+3-D) face recognition system with boosted performance. Performance of the proposed algorithm is evaluated on a large public database of range and portrait image pairs and found to perform quite well.

We develop algorithms that seek to assess the similarity of 3D faces, such that similar and dissimilar faces may be classified with high correlation relative to human perception of facial similarity. To obtain human facial similarity ratings, we conduct a subjective study, where a set of human subjects rate the similarity of pairs of faces. Such similarity scores are obtained from 12 subjects on 180 3D faces, with a total of 5490 pairs of similarity scores. We then extract Gabor features from automatically detected fiducial points on the range and texture images from the 3D face and demonstrate that these features correlate well with human judgements of similarity. Finally, we demonstrate the application of using such facial similarity ratings for scalable face recognition.

Automatic inspection of manufactured products with natural looking textures is a challenging task. Products such as tiles, textile, leather, and lumber project image textures that cannot be modeled as periodic or otherwise regular; therefore, a stochastic modeling of local intensity distribution is required. An inspection system to replace human inspectors should be flexible in detecting flaws such as scratches, cracks, and stains occurring in various shapes and sizes that have never been seen before. A computer vision algorithm is proposed in this paper that extracts local statistical features from grey-level texture images decomposed with wavelet frames into subbands of various orientations and scales. The local features extracted are second order statistics derived from grey-level co-occurrence matrices. Subsequently, a support vector machine (SVM) classifier is trained to learn a general description of normal texture from defect-free samples. This algorithm is implemented in LabVIEW and is capable of processing natural texture images in real-time.

We propose a novel technique to detect feature points from portrait and range representations of the face. In this technique, the appearance of each feature point is encoded using a set of Gabor wavelet responses extracted at multiple orientations and spatial frequencies. A vector of Gabor coefficients, called a jet, is computed at each pixel in the search window on a fiducial and compared with a set of jets, called a bunch, collected from a set of training data on the same type of fiducial. The desired feature point is located at the pixel whose jet is the most similar to the training bunch. This is the first time that Gabor wavelet responses were used to detect facial landmarks from range images. This method was tested on 1146 pairs of range and portrait images and high detection accuracies are achieved using a small number of training images. It is shown that co-localization using Gabor jets on range and portrait images resulted in better accuracy than using any single image modality. The obtained accuracies are competitive to that of other techniques in the literature.

In this paper a new framework for personal identity verification using 3-D geometry of the face is introduced. Initially, 3-D facial surfaces are represented by curves extracted from facial surfaces (facial curves). Two alternative facial curves are examined in this research: iso-depth and iso-geodesic curves. Iso-depth curves are produced by intersecting a facial surface with parallel planes perpendicular to the direction of gaze, at different depths from the nose tip. An Iso-geodesic curve is defined to be the locus of all points on the facial surface having the same geodesic distance from a given facial landmark (e.g. the nose tip). Once the facial curves are extracted, their characteristics are encoded by several features like the shape descriptors or polar Euclidean distances from the origin (nose tip). The final step is to verify or disapprove requests from users claiming the identity of registered individuals (gallery members) by comparing their features using Euclidean distance classifier or support vector machine (SVM). The performance results of the identity verification experiments are reported and a comparison is made between the two alternative curve-based facial surface representations.

In this paper, we present a novel identity verification system based on Gabor features extracted from range (3D) representations of faces. Multiple landmarks (fiducials) on a face are automatically detected using these Gabor features. Once the landmarks are identified, the Gabor features on all fiducials of a face are concatenated to form a feature vector for that particular face. Linear discriminant analysis (LDA) is used to reduce the dimensionality of the feature vector while maximizing the discrimination power. These novel features were tested on 1196 range images. The same features were also extracted from portrait images, and the accuracies of both modalities were compared. A superior verification accuracy was obtained using the range data, and a highly competitive accuracy to that of other techniques in the literature was also obtained for the portrait data.

Interest in face recognition systems has increased significantly due to the emergence of significant commercial opportunities in surveillance and security applications. In this paper we propose a novel technique to extract features from 3D face representations. In this technique, first the nose tip is automatically located on the range image, then the range data from a hexagonal region of interest around this landmark is decomposed using Barycentric wavelet kernels. The dimensionality of the extracted coefficients at each resolution level is reduced using principal component analysis (PCA). These new features are tested on 206 range images, and a high classification accuracy is achieved using a small number of features. The obtained accuracy is competitive to that of other techniques in literature.

As applications involving the capture of digital images become more ubiquitous – and at the same time more ambitious – there is a driving need for digital images of higher resolutions and quality. However, there is a limit to the spatial resolution that can be recorded by any digital device. Superresolution (SR) image reconstruction is the process of combining several low resolution images into a single higher resolution image. This allows the use of lower resolution (and thus lower cost) imaging systems than could otherwise be used for a given application. Due to these obvious benefits many SR reconstruction methods have been developed. We present an overview of existing SR methods and address the current need for an objective method to compare these techniques based on computational complexity and output quality.