Skip to main content
. Author manuscript; available in PMC: 2014 Jun 1.
Published in final edited form as: J Neural Eng. 2013 Apr 23;10(3):036015. doi: 10.1088/1741-2560/10/3/036015

Figure 2.

Figure 2

Pairs of dividing lines used to build orthogonal sets of spatial filters for extracting orthogonal movement components of 2D kinematic data. A) ‘CSP Cardinal’ set of orthogonal dividing lines. Continuous data was assigned to one of two classes across the dividing line based on the horizontal (filter set 1) or vertical (filter set 2) component of movement for generating spatial filters designed to extract the horizontal or vertical components of movement (respectively). B) ‘CSP Diagonal’ set of orthogonal dividing lines. Just as in panel A, continuous data was assigned to one of two classes based on which side of the dividing line it fell. The CSP Diagonal pair of dividing lines also generated two sets of spatial filters customized for detecting two orthogonal components of movement but along different set of orthogonal axes than the CSP Cardinal option.