Dr Chen Mao Davies

Interpreted the sensor data generated by sensor network systems to monitor the behaviours and conditions of vulnerable people in their homes to discern the lifestyle patterns and measures of well-being embedded in the data. This project is a collaboration among University of Dundee, British Telecom, and University of Liverpool. The research outcome has led to several journal and conference publications.

The research aims to utilise robust feature analysis to significantly improve the quality of interactive surface editing and manipulation of real scanned geometric models. The robust feature analysis extracts local geometric information such as salient curves, geometric patterns, self-similar shapes, geometric constraints and identifies mesh vertex/region correspondences. The feature guided interactive selection provides intuitive/interactive feedback and adaptive selection based on… Show more

The research aims to utilise robust feature analysis to significantly improve the quality of interactive surface editing and manipulation of real scanned geometric models. The robust feature analysis extracts local geometric information such as salient curves, geometric patterns, self-similar shapes, geometric constraints and identifies mesh vertex/region correspondences. The feature guided interactive selection provides intuitive/interactive feedback and adaptive selection based on pre-computed feature information. During feature guided interactive editing, both curve-based and surface-based shape editing can be performed at an interactive rate by utilising a uniform constrained optimisation framework. For example, editing operations exerted on a local region can be automatically propagated to similar shapes with pre-computed correspondences. Recognised geometric constraints can be preserved through an optimisation process. GPU based solvers can be utilised to improve the interactive editing performance. A prototype 3D shape editing tool is developed to allow more efficient and intuitive editing with improved quality and satisfying user experience.

This project is carried out in collaboration with Delcam Plc., a world leading provider of CAD/CAM software, providing a potential route to commercial in their products, including for example, dental CAD/CAM, ArtCAM, etc.
Show less

Thesis: An Investigation of Techniques for Sketch-based Human Modelling and Animation

Developed the methodologies and prototype system (Virtual Human Sketcher, VHS) for sketch-based virtual human modelling and animation. The research aimed to enable everyone who can draw, to “sketch-out” 3D virtual human body models and 2D/3D character animations, which are difficult to achieve by non-professional users.

VHS system/technology supports freehand drawing input and a “Stick… Show more

Thesis: An Investigation of Techniques for Sketch-based Human Modelling and Animation

Developed the methodologies and prototype system (Virtual Human Sketcher, VHS) for sketch-based virtual human modelling and animation. The research aimed to enable everyone who can draw, to “sketch-out” 3D virtual human body models and 2D/3D character animations, which are difficult to achieve by non-professional users.

VHS system/technology supports freehand drawing input and a “Stick Figure→Fleshing-out→Skin Mapping” modelling pipeline. Following this pipeline, a stick figure is drawn first to illustrate a figure pose, which is automatically reconstructed into 3D through a “Multi-layered Back-Front Ambiguity Clarifier”. It is then fleshed-out with freehand body contours. A “Creative Model-Method” (based on the elements of a human perception process) is developed for interpreting the body size, shape, and fat distribution of the sketched figure into a 3D human body surface model through graphical comparisons and generic model morphing. The generic model (generated from Virtual Human Project® CT cross-section images) is encapsulated with 3 layers: skeleton, fat tissue, and skin. It can be transformed sequentially through rigid morphing, fatness morphing, and surface matching to match the 2D figure sketch. The resulting 3D body model can be incrementally modified by sketching directly on the 3D model. In addition, this body surface can be mapped onto a series of posed stick figure to be interpolated as a 3D character animation. VHS supports sketch-based stick figure key framing and motion control as well as 2D/3D animations of multiple characters. VHS has been developed using User-Centred and Scenario-based Design approaches with continuous involvement of different users groups. The evaluation results suggest its usability and potentials to be used in a wide application field.
Show less