A Generative Model Of Perceived Glossiness Of Rough Surface

This paper focuses on human visual system’s perceptional glossiness of rough surfaces, andconstruct a generative model of perceptional glossiness of rough surfaces in mesoscale. The image ofrough surface is determined by three factors: the surface topology, light reflective properties, lightenvironment. We focuses on analyzing the influence of the environment and the surface topography topeople’s perception in this paper. In order to get the related data and the appearance of the target surfaceclosest to the stuff in the real scene, we choose a kind of noise surface which looks like real surfaces innature and it’s roughness can be parameterized, for the same reason, our rendering environment is alsomore complicated and more real than simple illuminant like point,line and surface illuminant. Althoughpeople established relevant physical models to describe the three factors, we can not portray theperceived dimension of the human visual system in the model’s parameter space. Much simulationrendering work has been done using our target surface and light environment based on physical model ofthe three factors, the rendering appearance of the surface are used in psychophysical experiments togather the perceptional gloss data of different rough surfaces under different conditions, then we analyzethe process of the rendering, firstly we make assumption the the rought surface is made of a largenumber of " micro-facet" structure, then we employ related optical geometry to explore the relationshipbetween the physical model and perceptional data, and render our target surface in math, besides, in thesimulation rendering process, we take the self-occlusion phenomenon among the " micro-facets" and thedistribution of the facets’ azimuth and slant into our consideration to build our generative model by thestatistics of the hight light percentage. The generative model can accurately predict the perceived glossof surfaces without rendering them in real, and it will bring application value in product design andmanufacturing, product quality control and subjective evaluation, seabed mineral exploration and otherfields.