Vehicle Detection on Unmanned Aerial Vehicle Images based on Saliency Region Detection

Li Wenhui,Qu Feng,Liu Peixun

Table 1 Details of test videos

Videos	Frame amount	Vehicle Amount	Resolution	Road condition	Weather condition	Largest vehicle amount
Test video 1	500	520	960 $V_{a} (i, j) = \| {wid}_{i} \times {hei}_{i} - {wid}_{j} \times {hei}_{j} \|$ 540	urban	cloudy	2
Test video 2	216	332	960 $D_{c} (i, j) = {‖c_{i} - c_{j}‖}^{2}$ 540	urban	sunny	3
Test video 3	500	1394	960 $D_{f} (i, j) = \frac{\max ? (f_{i}, f_{j})}{\min ? (f_{i}, f_{j})}$ 540	urban	sunny	7