Algorithm 1 Independent moving object detection

Input: ${\mathit{I}}_{\mathit{t}-1},{\mathit{I}}_{\mathit{t}},{\mathit{I}}_{\mathit{t}+1},{\mathit{H}}_{\mathit{t}-1,\mathit{t}}$ Output: ${\mathit{R}}_{\mathit{t}}$ 1: $\mathit{F}\mathit{P} \leftarrow \left[0 0 0 0 0 0\right]$ 2: $i=0$ 3: $\mathit{w}\mathit{h}\mathit{i}\mathit{l}\mathit{e} i<6:$ 4: $\mathit{F}\mathit{P}\left[i\right] \leftarrow number of feature points in {\mathit{I}}_{\mathit{t}}\left[i\right]$ 5: $if \mathit{F}\mathit{P}\left[i\right]<n:$ 6: $Find at least n-\mathit{F}\mathit{P}\left[i\right] feature points in {\mathit{I}}_{\mathit{t}}\left[i\right]$ 7: $\mathit{e}\mathit{n}\mathit{d}\mathit{i}\mathit{f}$ 8: $\mathit{e}\mathit{n}\mathit{d} \mathit{w}\mathit{h}\mathit{i}\mathit{l}\mathit{e}$ 9: ${\mathit{O}\mathit{F}}_{\mathit{t},\mathit{t}+1}\leftarrow optical flow between {\mathit{I}}_{\mathit{t}}\left[i\right] and {\mathit{I}}_{\mathit{t}+1}\left[i\right]$          ► Track background feature points 10: ${\mathit{H}}_{\mathit{t},\mathit{t}+1} \leftarrow estimate homography between {\mathit{I}}_{\mathit{t}}\left[i\right] and {\mathit{I}}_{\mathit{t}+1}\left[i\right] using \mathit{O}\mathit{F}$ ►Calculate homography matrix 11: ${\mathit{R}}_{\mathit{t}} \leftarrow average\left(\right[({\mathit{I}}_{\mathit{t}}-{\mathit{H}}_{\mathit{t}-1,\mathit{t}}·{\mathit{I}}_{\mathit{t}-1})+({\mathit{I}}_{\mathit{t}+1}-{\mathit{H}}_{\mathit{t},\mathit{t}+1}·{\mathit{I}}_{\mathit{t}})\left]\right)$ ►Background subtraction 12: ${\mathit{R}}_{\mathit{t}} \leftarrow Binarize\left(Dilation\right(MedianFilter\left({\mathit{R}}_{\mathit{t}}\right)\left)\right)$           ►Apply denoising filters and binarization