The main objective of the project is to render paints with the focus on high quality output so that the final rendering will highly correspond to the human perception of the real paints. The problem of appearance prediction of paint from direct measurements of Bi-directional Reflection Distribution Function (BRDF) is solved here.
We will focus mostly on modeling metallic and pearlescent paints. Rendering of coloured, metallic and pearlescent paints will be performed from BRDF data or eventually from Bi-directional Texture Function (BTF) data. We will consider both possibilities and choose the one, after considering benefits of each representation and additional statistical information about structure of metallic and pearlescent layer.
The goal of rendering is not to perform visualization in real time with large number of frames per second; how ever-endering software will try to take advantage of programmability, high performance and parallelism of modern graphics cards in processing of data to achieve interactive or eventually nearly interactive results.
We will propose a solution consisting of data acquisition from paint samples; new mathematical model based on multiple pigment layers; and new methods for storage and fast calculations on modern graphics processors (GP U).
We will study the undiscovered optical processes within the paints; we will formulate the relation between paint composition and final appearance, based on which the paint composition (pigments) will be determined.
Call for proposal
See other projects for this call