Size: 1677
Comment:
|
Size: 4320
Comment:
|
Deletions are marked like this. | Additions are marked like this. |
Line 3: | Line 3: |
Compiled by KD | |
Line 4: | Line 5: |
== Description == | |
Line 5: | Line 7: |
* The name of a landmark * The error for each image. No huge numbers of NaN |
RESIDUALS.TXT is a report created by [[residuals]] that looks for bad landmarks. |
Line 9: | Line 9: |
T has a map with it F flat landmark, no image landmark, resolution, lat, lon, radius |
A basic description of the information in RESIDUALS.TXT. * The post-fit residuals between predicted and actual control point positions in all images are captured in the residuals file * The post-fit residuals between predicted and actual control point difference between adjacent, overlapping maplets are captured in the residuals file * Outliers in both quantities are flagged automatically based on input threshold criteria A more in depth list explaining the information in the RESIDUALS.TXT file is shown below. * In between the dot rows: * The name of a landmark * T (landmark has a map with it) or F (flat landmark, no image) * Resolution * Lat of landmark * Wlon of landmark * Radius * Below the dot rows with the landmark name above them: * For each image: * 1st column - picture name * 2nd column - PX residual (# of pixels in image resolution) * 3rd column - LN residual (# of pixels in image resolution) * 4th column - magnitude of PX and LN residuals (# of pixels in image resolution) * 5th column - the estimate linear residual (km) * For each overlap: * 1st column - overlap name * 2nd column - X residual (km) * 3rd column - Y residual (km) * 4rd column - Z residual (km) * 5th column - estimate linear residual (km) * For each limb apperation: * 1st column - picture name * 2nd column - PX residual (# of pixels in image resolution) * 3rd column - LN residual (# of pixels in image resolution) * 4th column - magnitude of PX and LN residuals (# of pixels in image resolution) * 5th column - the estimate linear residual (km) * Below the images, overlaps, and limbs list for a landmark: * Landmark name * magnitude of RMS linear residual for landmark * Below the single dot row at the very bottom of the image: * Observations - how many IMAGEFILES, overlaps, and limbs for all landmarks * RMS RESIDUAL - overall RMS linear residual in meters. The user should note that residuals are the difference between observed landmark positions and those predicted by the solution (SPC). Observed landmark positions are exactly as the name suggests, positions of a landmark calculated by the code when it inputs pictures from the spacecraft of the object. Predicted landmark positions are those calculated by SPC when it inputs it generates a solution (map) from all the pictures collected. During iteration and refinement of the SPC solution, the position and attitude of the spacecraft are slightly changed over and over until the the maps match the images as best they can. At then end of each iteration SUMFILES are then updated for the new attitude and position. Thus, when the RMS between landmark residuals is as low as possible the maps match the images well, which means the updated SUMFILE is the new estimate for space craft position and attitude. |
Line 13: | Line 61: |
(List) Image name, pix residual, line residual (actual and solution should be), magnitude, km residual Mostly they are there for historical res |
|
Line 17: | Line 62: |
(Bottom) Meter resolution, difference between predicted and actual (Very bottom) Observations how many things RMS Residual (m) (vector residual) |
== Example of a RESIDUALS.TXT File == |
Line 50: | Line 89: |
Observation # = 360180 RMS Residual (m) = 0.78928860594347106 |
RESIDUALS.TXT
Compiled by KD
Description
RESIDUALS.TXT is a report created by residuals that looks for bad landmarks.
A basic description of the information in RESIDUALS.TXT.
- The post-fit residuals between predicted and actual control point positions in all images are captured in the residuals file
- The post-fit residuals between predicted and actual control point difference between adjacent, overlapping maplets are captured in the residuals file
- Outliers in both quantities are flagged automatically based on input threshold criteria
A more in depth list explaining the information in the RESIDUALS.TXT file is shown below.
- In between the dot rows:
- The name of a landmark
- T (landmark has a map with it) or F (flat landmark, no image)
- Resolution
- Lat of landmark
- Wlon of landmark
- Radius
- Below the dot rows with the landmark name above them:
- For each image:
- 1st column - picture name
- 2nd column - PX residual (# of pixels in image resolution)
- 3rd column - LN residual (# of pixels in image resolution)
- 4th column - magnitude of PX and LN residuals (# of pixels in image resolution)
- 5th column - the estimate linear residual (km)
- For each overlap:
- 1st column - overlap name
- 2nd column - X residual (km)
- 3rd column - Y residual (km)
- 4rd column - Z residual (km)
- 5th column - estimate linear residual (km)
- For each limb apperation:
- 1st column - picture name
- 2nd column - PX residual (# of pixels in image resolution)
- 3rd column - LN residual (# of pixels in image resolution)
- 4th column - magnitude of PX and LN residuals (# of pixels in image resolution)
- 5th column - the estimate linear residual (km)
- For each image:
- Below the images, overlaps, and limbs list for a landmark:
- Landmark name
- magnitude of RMS linear residual for landmark
- Below the single dot row at the very bottom of the image:
- Observations - how many IMAGEFILES, overlaps, and limbs for all landmarks
- RMS RESIDUAL - overall RMS linear residual in meters.
The user should note that residuals are the difference between observed landmark positions and those predicted by the solution (SPC).
Observed landmark positions are exactly as the name suggests, positions of a landmark calculated by the code when it inputs pictures from the spacecraft of the object.
Predicted landmark positions are those calculated by SPC when it inputs it generates a solution (map) from all the pictures collected.
During iteration and refinement of the SPC solution, the position and attitude of the spacecraft are slightly changed over and over until the the maps match the images as best they can. At then end of each iteration SUMFILES are then updated for the new attitude and position.
Thus, when the RMS between landmark residuals is as low as possible the maps match the images well, which means the updated SUMFILE is the new estimate for space craft position and attitude.
Example of a RESIDUALS.TXT File
.................................................. IA1_01 T 0.0200 56.5481 140.8124 13.6146 .................................................. A1_01 8.139 -6.853 10.640 0.01091 A2_05 -28.427 49.215 56.835 0.04250 B1_01 -3.517 52.347 52.465 0.04857 B1_03 -18.475 -57.091 60.006 0.05509 IA1_01 42.78 .................................................. IA1_03 T 0.0200 36.4771 136.8269 11.4593 .................................................. A1_03 -29.498 -12.161 31.906 0.04002 A2_03 -71.840 23.961 75.731 0.07142 A2_04 -3.885 -4.689 6.089 0.00643 A2_05 -34.754 95.249 101.392 0.10527 B1_01 64.895 -21.470 68.354 0.07337 B1_03 11.000 59.329 60.340 0.06832 B2_04 97.637 29.005 101.855 0.11002 B2_05 42.109 41.562 59.165 0.07455 IA1_03 75.45 ......... Observation # = 360180 RMS Residual (m) = 0.78928860594347106