Research projects of Evgueni Parilov

Numerical methods for nonlinear optics
The number of applications for materials exhibiting nonlinear multiphoton behavior has grown significantly in the last two decades, including but not limited to medical diagnostic tools, effective treatment for various cancer diseases, optical limiters, biological detectors, 3D microfabrication, fluorescent imaging system, and optical storage. We are working on developing robust numerical methods which accurately simulate the optical behavior of these materials under a high intensity laser irradiation. Our recent effort involved generating a new numerical framework for quantitative analysis of carrier dynamics of quantum dots.
project--

Normal maps with discontinuities
Standard bi- and trilinear interpolation of normal maps at subtexel level works fine for continuous normal fields, but may result in visible artifacts for surfaces with creases and dents. We develop real-time rendering technique which preserves the discontinuity curves and allows interactive blending of normal field with user-defined profile.
project--

Parametric synthesis
Accurate statistical description of two dimensional natural patterns is the basic building block for an automatic synthesis of such patterns. While hiding complicated undergoing processes which take part in forming such patterns, synthesis, based on such statistics, still produces reliable results. We are working on plausible models for synthesis of meso-scale human skin textures based on both local and global statistical descriptions.

Feature detection
Meso-scale human skin surface features, such as furrow networks and visible enlarged pores, have profound influence on skin appearance. We develop iterative computer vision techniques for detecting and analyzing such features from high-resolution digital images. The main difficulties are the inevitable presence of noise in skin images, the complexity of skin furrow network, and the shape irregularity of the features.

Human skin shading
Photorealistic rendering of human skin is still nontrivial task which requires applying many simulation techniques from different disciplines. Modeling intrinsic optical properties of the skin helps to predict skin shading color and allows to modify skin appearance in consistence with optical behavior of its real counterpart.

Closeup Skin Photography
Closeup photography
Macro photography is still one of the main tools in acquiring surface profiles of objects with highly detailed relief. Our focus is on photographing small patches of human skin. Highly diffuse skin reflection and close proximity to the subject from the camera lens make it very difficult to take large and sharp pictures of skin patches.

		Numerical methods for nonlinear optics The number of applications for materials exhibiting nonlinear multiphoton behavior has grown significantly in the last two decades, including but not limited to medical diagnostic tools, effective treatment for various cancer diseases, optical limiters, biological detectors, 3D microfabrication, fluorescent imaging system, and optical storage. We are working on developing robust numerical methods which accurately simulate the optical behavior of these materials under a high intensity laser irradiation. Our recent effort involved generating a new numerical framework for quantitative analysis of carrier dynamics of quantum dots. project--
		Normal maps with discontinuities Standard bi- and trilinear interpolation of normal maps at subtexel level works fine for continuous normal fields, but may result in visible artifacts for surfaces with creases and dents. We develop real-time rendering technique which preserves the discontinuity curves and allows interactive blending of normal field with user-defined profile. project--
		Parametric synthesis Accurate statistical description of two dimensional natural patterns is the basic building block for an automatic synthesis of such patterns. While hiding complicated undergoing processes which take part in forming such patterns, synthesis, based on such statistics, still produces reliable results. We are working on plausible models for synthesis of meso-scale human skin textures based on both local and global statistical descriptions.
		Feature detection Meso-scale human skin surface features, such as furrow networks and visible enlarged pores, have profound influence on skin appearance. We develop iterative computer vision techniques for detecting and analyzing such features from high-resolution digital images. The main difficulties are the inevitable presence of noise in skin images, the complexity of skin furrow network, and the shape irregularity of the features.
		Human skin shading Photorealistic rendering of human skin is still nontrivial task which requires applying many simulation techniques from different disciplines. Modeling intrinsic optical properties of the skin helps to predict skin shading color and allows to modify skin appearance in consistence with optical behavior of its real counterpart.
		Closeup photography Macro photography is still one of the main tools in acquiring surface profiles of objects with highly detailed relief. Our focus is on photographing small patches of human skin. Highly diffuse skin reflection and close proximity to the subject from the camera lens make it very difficult to take large and sharp pictures of skin patches.