Nice work. The subdivision into cells and encoding into GPU textures is strongly reminiscent of the 2008 paper "Random-Access Rendering of General Vector Graph

Nice work. The subdivision into cells and encoding into GPU textures is strongly reminiscent of the 2008 paper "Random-Access Rendering of General Vector Graphics" by Nehab and Hoppe [1]. Rotating the line samples is a rather interesting idea. Seems like that would converge to the equivalent of a convolution with a radial tent filter. [1] http://w3.impa.br/~diego/publications/NehHop08.pdf http://w3.impa.br/~diego/publications/NehHop08.pdf