Abstract
Dirty floors and walls are commonly viewed in the place where many people walk over. Intensity of the grime increases in proportion to the number of people passing through the place. And pattern of the grime changes by the walking speed of the people. Therefore, we increase reality of CG images of floors and walls by the following method. First, motion pass and speed of each person is computed by crowd simulation. Then foot patterns are allocated along the pass after modified according to the speed. If a person walks near a wall, contact to the wall is evaluated. If contact, one grime pattern selected from data base is overwritten on the wall at height decided with the probability distribution function that is measured from sample walls.