Optimum Pipeline for Visual Terrain Classification Using Improved Bag of Visual Words and Fusion Methods

Figure 6

The testing time of 1200 terrain images (256 pixels × 256 pixels) for different encoding methods with varying codebook size on the DS1 dataset for SIFT and DSIFT descriptors. (a) SIFT (sparse descriptor); (b) DSIFT (dense descriptor). The numbers in -axis outside (inside) the parenthesis indicate the codebook size for activation (difference)-based coding methods.

(a)

(b)