Real Scenario Parameter Extraction by Using PSO Rendering in the Loop

Realistic image synthesis is highly dependent on rendering algorithm and optical properties of scenario objects. The goal of this study, which is still in progress, proposes a methodology to obtain some illumination parameters of a real scenario represented by an acquired image, and to use parameters for a virtual scenario with the same object as the original. The proposed methodology consists, first, in acquiring an image of the working scenario, and by using a particle swarm optimization renderer in the loop, rendering images that gradually approximate to the real acquired image, by some virtual scenario parameter modification based on the PSO (Particle Swarm Optimization) search. We call it “Rendering in the loop”. Finally we use the obtained parameters to render a better resolution image and see the result.