Hole Targeting Set-up for MSI-Edge » History » Version 3
Amber Herold, 06/22/2010 02:37 PM
1 | 1 | Amber Herold | h1. Hole Targeting Set-up for MSI-Edge |
---|---|---|---|
2 | |||
3 | |||
4 | We can use any image that is already loaded into Hole Targeting to adjust these parameters. The values in this node vary so much from grid to grid that the best way to find these parameters is by going through a step-by-step trial and error process. The parameters not mentioned here can be left at their default values. To proceed from one step of the hole targeting process to another, simply proceed from top to bottom through the display selection buttons in the image control panel. The display settings associated with each display selection are the locations where the Hole Targeting parameters can be adjusted. To see the final acquisition and focus targets, enable only the Original, acquisition, and focus display selections. |
||
5 | |||
6 | |||
7 | |||
8 | 3 | Amber Herold | 1. Leginon/Hole Targeting/Original> |
9 | 1 | Amber Herold | |
10 | |||
11 | Use this setting to load a test image if needed. |
||
12 | |||
13 | |||
14 | |||
15 | 3 | Amber Herold | * Select the Image Display Selection before opening the Image Display Settings window. |
16 | 1 | Amber Herold | |
17 | |||
18 | 3 | Amber Herold | * Browse the rawdata directory of the current session for an example image for testing. |
19 | 1 | Amber Herold | |
20 | |||
21 | 3 | Amber Herold | * Load the selected file and exit Image Display Settings window. |
22 | 1 | Amber Herold | |
23 | |||
24 | |||
25 | |||
26 | 3 | Amber Herold | 2. (This step does not apply if you are using [[JAHC finder nodes set-up|MSI-T]]) |
27 | 1 | Amber Herold | |
28 | |||
29 | Leginon/Hole Targeting/Edge> |
||
30 | |||
31 | |||
32 | A good combination of parameters show the edges of holes well with minimal thresholded "edge" from random features. |
||
33 | |||
34 | 3 | Amber Herold | * Select the Edge Display Selection before opening the Edge Display Settings window. |
35 | 1 | Amber Herold | |
36 | 3 | Amber Herold | * Choose an edge detection Threshold and press "Test". "Test" should be done every time a parameter is changed to check its effect. |
37 | 1 | Amber Herold | |
38 | 3 | Amber Herold | * Adding a proper low pass filter can sometimes reduce the fragmented appearance of the edge image. |
39 | 1 | Amber Herold | |
40 | 3 | Amber Herold | * Increase the Threshold if too many edges are shown in the edge image. |
41 | 1 | Amber Herold | |
42 | 3 | Amber Herold | * Use the ruler tool to measure the average inner and outer hole diameters for preparation on template correlation. |
43 | 1 | Amber Herold | |
44 | |||
45 | 3 | Amber Herold | 3. (This step is different if you are using <link linkend="MSIT_tmplt_cor">MSI-T</link>) |
46 | 1 | Amber Herold | |
47 | |||
48 | Leginon/Hole Targeting/Template> |
||
49 | |||
50 | A good template gives a sharp correlation at the center of the holes. |
||
51 | |||
52 | 3 | Amber Herold | * Select the Template Display Selection before opening the Template Display Settings window. |
53 | 1 | Amber Herold | |
54 | |||
55 | 3 | Amber Herold | * Click Add (Rings) and enter the inner and outer hole diameters that were measured in the previous section. (Only one set of hole diameters should be in the list. If a set already exists, click Edit instead of entering the measure hole diameters.) |
56 | 1 | Amber Herold | |
57 | |||
58 | 3 | Amber Herold | * Choose a correlation method. "cross" correlation is the typical for high contrast images. "phase" correlation works better with weak hole edges (i.e., thick ice) but should be used with a low pass filter. For example, increasing the filter pixel size can eliminating scattered peaks. |
59 | 1 | Amber Herold | |
60 | |||
61 | 3 | Amber Herold | * "Test" these parameters and adjust, normally, the template diameters for sharpest correlation peaks at the center of the holes and exit the setting window. |
62 | 1 | Amber Herold | |
63 | |||
64 | |||
65 | 3 | Amber Herold | 4. Leginon/Hole Targeting/Threshold> |
66 | 1 | Amber Herold | |
67 | A good correlation threshold leaves only small blobs of the correlation peaks from the holes. Since the peaks from holes with ice are usually weaker but more important to catch, it is o.k. to leave some but not too many noise peaks. This parameter also tends to change during an experiment. The value is in the unit of number of standard deviation above the mean. |
||
68 | |||
69 | 3 | Amber Herold | * Select the Threshold Display Selection before opening the Threshold Display Settings window. |
70 | 1 | Amber Herold | |
71 | 3 | Amber Herold | * Enter the Threshold value determined in the previous section and press Test. |
72 | 1 | Amber Herold | |
73 | |||
74 | 3 | Amber Herold | 5. Leginon/Hole Targeting/Blobs> |
75 | 1 | Amber Herold | |
76 | If the threshold level from the last step is too low, there will be a lot of big blobs, making the process slow. The parameters set in this node reasonably limit the total number of blobs. |
||
77 | |||
78 | 3 | Amber Herold | * Select the Blob Display Section before opening the Blob Display Settings window. |
79 | 1 | Amber Herold | |
80 | 3 | Amber Herold | * "Border" should be set to at least the radius of the holes in order to reduce the border effects. |
81 | 1 | Amber Herold | |
82 | 3 | Amber Herold | * "Maximum number of blobs" is typically 300 (although it is sometimes larger if you expect more holes). |
83 | 1 | Amber Herold | |
84 | 3 | Amber Herold | * "Maximum size of blobs" should be adjusted until mostly hole centers, not junk spots, are shown. |
85 | 1 | Amber Herold | |
86 | 3 | Amber Herold | * "Test" and adjust the parameters in the last three steps (2-4) to acheive the best results. |
87 | 1 | Amber Herold | |
88 | 3 | Amber Herold | * To help define the lattice, measure the distance between hole centers with the ruler tool. |
89 | 1 | Amber Herold | |
90 | |||
91 | 3 | Amber Herold | 6. Leginon/Hole Targeting/Lattice> |
92 | 1 | Amber Herold | |
93 | The lattice of the holes is not perfect; therefore, a decent tolerance is necessary for a good fit to this imperfect lattice. If only a few blobs were centered in the last step, then this step is not likely to work. |
||
94 | |||
95 | 3 | Amber Herold | * Select the Lattice Display Section before opening the Lattice Display Settings window. |
96 | 1 | Amber Herold | |
97 | 3 | Amber Herold | * Enter the Spacing from the measurement in the last step. |
98 | 1 | Amber Herold | |
99 | 3 | Amber Herold | * Typical Tolerance = 0.1 (i.e., 10% error in the spacing is accepted) |
100 | 1 | Amber Herold | |
101 | 3 | Amber Herold | * Hole Stats Radius = ~ 3/4 of the hole radius. |
102 | 1 | Amber Herold | |
103 | 3 | Amber Herold | * Zero thickness = the intensity measurement from an empty hole at the same preset (the sq preset). |
104 | 1 | Amber Herold | |
105 | 3 | Amber Herold | * "Test" and examine the "mean thickness" and "standard deviation of thickness" in each hole on the lattice. |
106 | 1 | Amber Herold | |
107 | 3 | Amber Herold | 7. Leginon/Hole Targeting/acquisition> |
108 | 1 | Amber Herold | |
109 | The settings here depend on the required ice thickness and on the "Zero thickness" setting in the last step. At a minimum, these settings can be used to rule out holes that are empty and that have massive ice contaminant. |
||
110 | |||
111 | 3 | Amber Herold | * Select the Original and acquisition Display Sections before opening the acquisition Display Settings window. |
112 | 1 | Amber Herold | |
113 | 3 | Amber Herold | * Min. mean = ~ 0.05 measured by ln(I0/I) |
114 | 1 | Amber Herold | |
115 | 3 | Amber Herold | * Max. mean = ~ 0.3 |
116 | 1 | Amber Herold | |
117 | 3 | Amber Herold | * Max. stdev = ~ 0.2 |
118 | 1 | Amber Herold | |
119 | 3 | Amber Herold | * Focus hole selection = Good hole ( or "Any Hole" if hole targets are not plentyful. With "Any Hole" option,and a bad hole may be targeted that is hard to focus such as very thick ice. Alternatively, "Off" to skip the Z height correction step at each square ) |
120 | 1 | Amber Herold | |
121 | 3 | Amber Herold | * Use target template = no |
122 | 1 | Amber Herold | |
123 | |||
124 | 3 | Amber Herold | 8. Leginon/Hole Targeting/focus> |
125 | 1 | Amber Herold | |
126 | 3 | Amber Herold | The only relevant focus setting ("Focus hole selection") in Hole Targeting is in relation to Z Height correction. The Leginon/Hole Targeting/focus Display Settings window = Leginon/Hole Targeting/acquisition Display Settings window. This means that the focus settings were set in the previous section and do not need to be revisited here. To see the focus targets, enable the Original and focus (and optionally the acquisition) display selections in the image control panel. |
127 | 1 | Amber Herold | |
128 | |||
129 | |||
130 | 3 | Amber Herold | 9. Leginon/Hole Targeting/Settings> Activate automatic hole finding for future images received. |
131 | 1 | Amber Herold | |
132 | |||
133 | Skip automatic hole finding = no |
||
134 | 2 | Amber Herold | |
135 | |||
136 | ______ |
||
137 | |||
138 | [[Summary of this application|< Summary of this application]] | [[Exposure Targeting Set-up for MSI-Edge|Exposure Targeting Set-up for MSI-Edge >]] |
||
139 | |||
140 | ______ |