Skip to main content
. 2018 Sep 3;74(Pt 9):877–894. doi: 10.1107/S2059798318009191

Figure 2.

Figure 2

Automatic quadrant alignment by rotational autocorrelation from thermolysin data. (a) Composite maximum images from run 22, run 14 and runs 13–22. The red arcs and dotted extensions have been placed to show that the virtual powder rings are not circular. (b) Rotational autocorrelations of quadrant 2. The x and y axes represent the incremental shifts in the quadrant position used to calculate the autocorrelation, with the center coordinate (0, 0) representing no shift. The heat map is colored by the rotational autocorrelation of the quadrant with itself, rotated 45° around the beam center. Heat maps are all colored on the same scale (see color bar). The maximum value is marked with a green dot. (c) Autocorrelation map using multiple rotation angles. Each point is the maximum CC value found when rotating the panel in steps from 20 to 70° in 2.5° increments. (d) Diagram illustrating rotational autocorrelation. The quadrant is rotated 45° and a CC is computed between the pixel values in the overlapping areas. This is repeated for each of the grid points after translating the quadrant. (e) The maximum composite pattern from runs 13–22 after applying the x and y shifts for each quadrant that maximize the rotational autocorrelation.