Save object rendering costs during pan by re-using previous data

At present, when map objects are being rendered, at each individual frame change of the camera data, the vertices and location of map objects such as polygons and circles is recalculated, a potentially expensive operation. It may be possible to re-use a previous rendering of an object during animation frames to avoid this re-calculation, by making assumptions about panning on the projection and rendering to a slightly larger surface to avoid some clipping.