Algorithm 1.

The proposed integrated segmentation algorithm incorporating the variant moments and the measures of optic flow, flow’s magnitude and phase of each pixel (m_ij, θ_ij).

1:	Function Segmentation(I)
2:	{An image I of N × M pixel intensities}
3:	fori = 1 to N do
4:	forj = 1 to M do
5:	Obtain a window: $$W_{\mathit{ij}}=I\left[i-\frac{w}{2}\dots i+\frac{w}{2},j-\frac{h}{2}\dots j+\frac{h}{2}\right]\text{of} w\times h \text{neighbours of} I[i,j]$$
6:	Obtain pixel features: $$F_{\mathit{ij}}\leftarrow \left[\begin{array}{cc}\stackrel{x,y-\mathit{coordinates}}{\overbrace{x(i),y(j),}}& \end{array}\begin{array}{cc}\underset{\mathit{variant}-\mathit{moments}}{\underbrace{A(W_{\mathit{ij}}),L_x(W_{\mathit{ij}}),L_y(W_{\mathit{ij}}),P_x(W_{\mathit{ij}}),P_y(W_{\mathit{ij}}),}}& \stackrel{\mathit{optic}-\mathit{flow}}{\overbrace{m_{\mathit{ij}},{\theta }_{\mathit{ij}}}}\end{array}\right]$$
7:	end for
8:	end for
9:	class-id = k-means(F,k,w)
10:	returnclass-id