Naoufel Werghi

Finding Surface Correspondance for Object Recognition and Registration Using Pairwise Geometric Histograms

Pairwise geometric histograms have been demonstrated as an effective descriptor of arbitrary 2-dimensional shape which enable robust and efficient object recognition in complex scenes. In this paper we describe how the approach can be extended to allow the representation and classification of arbitrary 2 1/2- and 3-dimensional surface shape. This novel representation can be used in important vision tasks such as the recognition of objects with complex free-form surfaces and the registration of surfaces for building 3-dimensional models from multiple views. We apply this new representation to both of these tasks and present some promising results.

Object reconstruction by incorporating geometric constraints in reverse engineering

This paper deals with the constrained reconstruction of 3D geometric models of objects from range data. It describes a new technique of global shape improvement based upon feature positions and geometric constraints. It suggests a general incremental framework whereby constraints can be added and integrated in the model reconstruction process, resulting in an optimal trade-off between minimization of the shape fitting error and the constraint tolerances. After defining sets of constraints for planar and special case quadric surface classes based on feature coincidence, position and shape, the paper shows through application on synthetic model that our scheme is well behaved. The approach is then validated through experiments on different real parts. This work is the first to give such a large framework for the integration of geometric relationships in object modelling. The technique is expected to have a great impact in reverse engineering applications and manufactured object modelling where the majority of parts are designed with intended feature relationships.

A discrete Reeb graph approach for the segmentation of human body scans

Segmentation of 3D human body (HB) scan is a very challenging problem in applications exploiting human scan data. To tackle this problem, we propose a topological approach based on discrete Reeb graph (DRG) which is an extension of the classical Reeb graph to unorganized cloud of 3D points. The essence of the approach is detecting critical nodes in the DRG thus permitting the extraction of branches that represent the body parts. Because the human body shape representation is built upon global topological features that are preserved so long as the whole structure of the human body does not change, our approach is quite robust against noise, holes, irregular sampling, moderate reference change and posture variation. Experimental results performed on real scan data demonstrate the validity of our method.

Matching 3D face scans using interest points and local histogram descriptors

In this work, we propose and experiment an original solution to 3D face recognition that supports face matching also in the case of probe scans with missing parts. In the proposed approach, distinguishing traits of the face are captured by first extracting 3D keypoints of the scan and then measuring how the face surface changes in the keypoints neighborhood using local shape descriptors. In particular: 3D keypoints detection relies on the adaptation to the case of 3D faces of the meshDOG algorithm that has been demonstrated to be effective for 3D keypoints extraction from generic objects; as 3D local descriptors we used the HOG descriptor and also proposed two alternative solutions that develop, respectively, on the histogram of orientations and the geometric histogram descriptors. Face similarity is evaluated by comparing local shape descriptors across inlier pairs of matching keypoints between probe and gallery scans. The face recognition accuracy of the approach has been first experimented on the difficult probes included in the new 2D/3D Florence face dataset that has been recently collected and released at the University of Firenze, and on the Binghamton University 3D facial expression dataset. Then, a comprehensive comparative evaluation has been performed on the Bosphorus, Gavab and UND/FRGC v2.0 databases, where competitive results with respect to existing solutions for 3D face biometrics have been obtained. Graphical abstractDisplay Omitted Highlights3D face recognition approach deployable in real non-cooperative contexts of use.Fully-3D approach, based on keypoints detection, description and matching.MeshDOG keypoints detector combined with the multi-ring GH descriptor.RANSAC algorithm included for outlier removal from matching keypoints.State of the art accuracy for recognizing 3D scans with missing parts.

A functional-based segmentation of human body scans in arbitrary postures

This paper presents a general framework that aims to address the task of segmenting three-dimensional (3-D) scan data representing the human form into subsets which correspond to functional human body parts. Such a task is challenging due to the articulated and deformable nature of the human body. A salient feature of this framework is that it is able to cope with various body postures and is in addition robust to noise, holes, irregular sampling and rigid transformations. Although whole human body scanners are now capable of routinely capturing the shape of the whole body in machine readable format, they have not yet realized their potential to provide automatic extraction of key body measurements. Automated production of anthropometric databases is a prerequisite to satisfying the needs of certain industrial sectors (e.g., the clothing industry). This implies that in order to extract specific measurements of interest, whole body 3-D scan data must be segmented by machine into subsets corresponding to functional human body parts. However, previously reported attempts at automating the segmentation process suffer from various limitations, such as being restricted to a standard specific posture and being vulnerable to scan data artifacts. Our human body segmentation algorithm advances the state of the art to overcome the above limitations and we present experimental results obtained using both real and synthetic data that confirm the validity, effectiveness, and robustness of our approach.

A low-cost range finder using a visually located, structured light source

In this paper we show how the cost of a structured light, range finding system can be substantially reduced by visually tracking the structured light source. To be able to recover range measurements using a structured light range finder the relative positions of the structured light source and light sensor must be known. This is typically achieved by carefully controlling the position of at least one of these components using expensive mechanical actuators. Instead, we propose that little or no control is placed on the positioning of these components and that the position of the structured light source is determined using visual feedback. A low-cost prototype system employing this principle is presented.

The Mesh-LBP: A Framework for Extracting Local Binary Patterns From Discrete Manifolds

In this paper, we present a novel and original framework, which we dubbed mesh-local binary pattern (LBP), for computing local binary-like-patterns on a triangular-mesh manifold. This framework can be adapted to all the LBP variants employed in 2D image analysis. As such, it allows extending the related techniques to mesh surfaces. After describing the foundations, the construction and the main features of the mesh-LBP, we derive its possible variants and show how they can extend most of the 2D-LBP variants to the mesh manifold. In the experiments, we give evidence of the presence of the uniformity aspect in the mesh-LBP, similar to the one noticed in the 2D-LBP. We also report repeatability experiments that confirm, in particular, the rotation-invariance of mesh-LBP descriptors. Furthermore, we analyze the potential of mesh-LBP for the task of 3D texture classification of triangular-mesh surfaces collected from public data sets. Comparison with state-of-the-art surface descriptors, as well as with 2D-LBP counterparts applied on depth images, also evidences the effectiveness of the proposed framework. Finally, we illustrate the robustness of the mesh-LBP with respect to the class of mesh irregularity typical to 3D surface-digitizer scans.

Topological segmentation of discrete human body shapes in various postures based on geodesic distance

This paper extends our previous Reeb graph approach based on a new Morse function, namely geodesic distance, to segment whole body scan data into primary body parts in various postures. Because of the bending invariance of geodesic distance, the resulting Reeb graph can remain stable in a large range of postures. Consequently, the approach is capable of segmenting data within the posture range. The application of geodesic distance also brings the independence of coordinate frame selection. We present a number of experiments conducted on both real body 3D scan samples and simulated datasets to demonstrate the validity of the approach.

Modelling Objects having Quadric Surfaces Incorporating Geometric cCnstraints

This paper deals with the constrained shape reconstruction of objects having quadric patches. The incorporation of geometric constraints in object reconstruction was used first by Porrill [10]. His approach combined the Kalman filter equations with linearized constraint equations. This technique was improved by De Geeter et al [5] to reduce the effects of linearization error. The nature and the specificity of this technique make it limited in scope and application.

Selecting stable keypoints and local descriptors for person identification using 3D face scans

Abstract3D face identification based on the detection and comparison of keypoints of the face is a promising solution to extend face recognition approaches to the case of 3D scans with occlusions and missing parts. In fact, approaches that perform sparse keypoints matching can naturally allow for partial face comparison. However, such methods typically use a large number of keypoints, locally described by high-dimensional feature vectors: This, combined with the combinatorial number of keypoint comparisons required to match two face scans, results in a high computational cost that does not scale well with large datasets. Motivated by these considerations, in this paper, we present a 3D face recognition approach based on the meshDOG keypoints detector and local GH descriptor, and propose original solutions to improve keypoints stability and select the most effective features from the local descriptors. Experiments have been performed to assess the validity of the proposed optimizations for stable keypoints detection and feature selection. Recognition accuracy has been evaluated on the Bosphorus database, showing competitive results with respect to existing 3D face identification solutions based on 3D keypoints.