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

Table 2

Comparison of different fusion methods for the coding methods for SIFT on DS1. The average classification accuracy (%) is set as the evaluation index.

Methods	SA	LSA	LLC	SVC	SVC_K	VLAD	VLAD_K	FV
SIFT	72.17	76.13	76.05	76.97	79.23	76.50	80.13	82.38
GIST	73.55
Averaging	83.17	86.12	85.60	85.20	87.70	85.72	86.45	88.50
Product	75.14	81.00	80.36	67.36	75.78	67.50	77.28	74.28