Anthropometric accuracy of three-dimensional average faces compared to conventional facial measurements

This study aimed to evaluate and compare the accuracy of average faces constructed by different methods. Original three-dimensional facial images of 26 adults in Chinese ethnicity were imported into Di3DView and MorphAnalyser for image processing. Six average faces (Ave_D15, Ave_D24, Ave_MG15, Ave_MG24, Ave_MO15, Ave_MO24) were constructed using “surface-based registration” method with different number of landmarks and template meshes. Topographic analysis was performed, and the accuracy of six average faces was assessed by linear and angular parameters in correspondence with arithmetic means calculated from individual original images. Among the six average faces constructed by the two systems, Ave_MG15 had the highest accuracy in comparison with the conventional method, while Ave_D15 had the least accuracy. Other average faces were comparable regarding the number of discrepant parameters with clinical significance. However, marginal and non-registered areas were the most inaccurate regions using Di3DView. For MorphAnalyser, the type of template mesh had an effect on the accuracy of the final 3D average face, but additional landmarks did not improve the accuracy. This study highlights the importance of validating software packages and determining the degree of accuracy, as well as the variables which may affect the result.

Landmark-based homologous multi-point warping approach to 3D facial recognition using multiple datasets

Over the years, neuroscientists and psychophysicists have been asking whether data acquisition for facial analysis should be performed holistically or with local feature analysis. This has led to various advanced methods of face recognition being proposed, and especially techniques using facial landmarks. The current facial landmark methods in 3D involve a mathematically complex and time-consuming workflow involving semi-landmark sliding tasks. This paper proposes a homologous multi-point warping for 3D facial landmarking, which is verified experimentally on each of the target objects in a given dataset using 500 landmarks (16 anatomical fixed points and 484 sliding semi-landmarks). This is achieved by building a template mesh as a reference object and applying this template to each of the targets in three datasets using an artificial deformation approach. The semi-landmarks are subjected to sliding along tangents to the curves or surfaces until the bending energy between a template and a target form is minimal. The results indicate that our method can be used to investigate shape variation for multiple datasets when implemented on three databases (Stirling, FRGC and Bosphorus).

3-Dimensional facial expression recognition in human using multi-points warping

Background Expression in H-sapiens plays a remarkable role when it comes to social communication. The identification of this expression by human beings is relatively easy and accurate. However, achieving the same result in 3D by machine remains a challenge in computer vision. This is due to the current challenges facing facial data acquisition in 3D; such as lack of homology and complex mathematical analysis for facial point digitization. This study proposes facial expression recognition in human with the application of Multi-points Warping for 3D facial landmark by building a template mesh as a reference object. This template mesh is thereby applied to each of the target mesh on Stirling/ESRC and Bosphorus datasets. The semi-landmarks are allowed to slide along tangents to the curves and surfaces until the bending energy between a template and a target form is minimal and localization error is assessed using Procrustes ANOVA. By using Principal Component Analysis (PCA) for feature selection, classification is done using Linear Discriminant Analysis (LDA). Result The localization error is validated on the two datasets with superior performance over the state-of-the-art methods and variation in the expression is visualized using Principal Components (PCs). The deformations show various expression regions in the faces. The results indicate that Sad expression has the lowest recognition accuracy on both datasets. The classifier achieved a recognition accuracy of 99.58 and 99.32% on Stirling/ESRC and Bosphorus, respectively. Conclusion The results demonstrate that the method is robust and in agreement with the state-of-the-art results.

A parametric morphing method for generating structured meshes for marine free surface flow applications with plane symmetry

This article introduces a method for morphing a parametric, rectangular and structured 3D template mesh such that it embodies: (1) a given (half) hull surface properly aligned with a pre-defined air-water interface zone, (2) a high-quality boundary layer region near the hull, and (3) a smooth transition of the mesh to the template mesh structure away from the hull. The performance of the proposed method is successfully validated on three widely studied benchmark cases, namely the parabolic Wigley hull, the Series 60 hull and the DTMB 5415 combatant hull. The proposed automated mesh generation method is useful in shape optimization for computational fluid dynamics, among others. Highlights A new method for creating structured meshes for marine CFD applications is presented. The proposed method is parametric and thus fully programmable. The method includes an automatic resolution of CAD geometry errors. The resulting meshes are strictly orthogonal near the inter-phase zone and the hull. The main application of the proposed method is in CFD-based shape optimization.

Comparison of Displacement-Based and Force-Based Mapped Meshing

The finite element (FE) method is a powerful tool for the study of biomechanics. One of the limiting factors in transitioning this tool into the clinic is the time required to generate high quality meshes for analysis. Previously, we developed a mapped meshing technique that utilized force control and a finite element solver to warp a template mesh onto subject specific surfaces. This paper describes a displacement based method that directly warps the template mesh onto subject specific surfaces using distance as the driving measure for the deformable registration. The resulting meshes were evaluated for mesh quality and compared to the force based method. An initial evaluation was performed using a mathematical phantom. The algorithm was then applied to generate meshes for the phalanx bones of the human hand. The algorithm successfully mapped the template bone to all of the bony surfaces, with the exception of the distal phalanx bone. In this one case, significant differences existed between the geometries of the template mesh and the distal phalanx. Further refinement of the algorithm may allow the algorithm to successfully generate meshes even in the presence of large geometric shape differences.