Figure 1. Sample grid around detected key-point. 8×8 highlighted pixels are used to create descriptor. Post navigationPrevious post: Planar Object Recognition In An Augmented Reality Application On Mobile Devices