This kind of Field of View algorithm is integral to any roguelike and there's a variety of methods for achieving the results [0]. I'd be interested in how this algorithm holds up against the usual raycasting or shadow casting. My intuition is that it'd be slower (especially with a 1024px shadow map resolution) due to it being based on pixels rather than grid values.

[0] http://roguebasin.roguelikedevelopment.org/index.php/Field_o...