Differences between revisions 11 and 47 (spanning 36 versions)

residuals

Compiled by KD

Description

The routine residuals tests the landmarks to look for problems and gives reports about how well the landmarks were processed. In other words, residuals shows how well SPC performed. The reports that residuals outputs are listed in the Output section.

Input - files

LMKFILES
NOMINALS
SUMFILES
IMAGEFILES
PICTLIST.TXT - list of picture names generated by make_sumfiles
LMRKLIST.TXT

Optional Files

PICTLISTS.TXT - user-generated list of picture names, if this file exists, it is used instead of PICTLIST.TXT

Input - stdin

enter 3 values (px,km,km)
1. The value of the pixel residual limit. Images and maplets are flagged for a pixel residual limit greater than this.
2. The value of the linear residual beyond which images and maplets are flagged.
3. The max scale (km/px) that sets the scale for the histogram at the end of PICINFO.TXT.
input operation list

     10, 0.01, 0.01

Output

MAPINFO.TXT
PICINFO.TXT
LIMINFO.TXT
RESIDUALS.TXT
check.txt - List of landmarks and overlaps for landmarks whose linear residual is greater than the user-specified limit.
EMPTY.TXT - lithos seed file for batch deletion of landmarks which are not contained in any pictures or limbs.
FLATLIST.TXT - List of flat maps (landmarks containing no topography).
LMKVECS.TXT - Landmark vectors for every landmark contained in LMRKLIST.TXT.
MAPCHK.TXT - List of maplets whose difference between predicted and observed pixel/line locations in attached images is greater than a user-specified limit, or who have two or fewer overlaps attached to them.
MAPRES.TXT - Maplet resolution information for maplets contained in at least one picture or limb.
New_Limbs.in - lithos seed file to attach map to limbs for every landmark contained in LMRKLIST.TXT.
no_update.txt - List of landmarks whose associated pictures are not listed by increasing mission time in the .LMK file.
PRUNE.TXT - List of landmarks for which the number of pictures in which they are contained exceeds the default limit of 500, or the user-specified limit, PRNLM, set within INIT_LITHOS.TXT.
RANGES_SOLVED.TXT - Date and range information for each picture listed in PICTLISTS.TXT or PICTLIST.TXT.
veto.txt - lithos seed file to detach from landmarks and limbs a map with a linear residual greater than the user-specified limit.

Example of MAPINFO.TXT

Example of a single landmark in RESIDUALS.TXT

What a Bad & Fixed Image Looks Like

Qualitative Checks

The routine residuals is not the only way to check how well SPC performed. One can simply observe the maplets generated. The information below is what to look for when visually inspecting a maplet to judge its quality.

Qualitative checks are important to verify that the quantitative error estimates are believable
Render a single maplet, or a DEM synthesized by a collection of maplets, at the same geometry & illumination conditions as the images themselves
- Do all the features in the images appear in the DEM ?
- Are the smallest discernible features in the images, e.g., boulders, craters etc. visible in the DEM as well ?
- Is the relative albedo solution such that the relative brightness of adjacent image features matches that of the DEM ?
- Are the heights solution such that the length and overall appearance of the shadows in the DEM matches that of the image ?

How Uncertainties are Introduced in the Data

Main types of uncertainties:
- S/C trajectory, camera pointing, image timestamp
  - Manifest themselves mostly in the projection of image template onto the surface. Corrected by global geometry solution
- Image noise, artifacts, smear, overall image quality
  - Manifest themselves in predicted image template brightness and in its fit to the extracted image brightness
- Photometric model and reflectance function models
  - Show up in slopes & heights integration
- Poor choice of a-priori parameters & data weights
  - Evident at end of each estimation step; data won’t fit well
Often above contributions are correlated; individual contributors may not be separated until many processing steps have been taken

CategoryPrograms

-  ⇤ ← Revision 11 as of 2015-05-15 11:35:50 → 
  Size: 2804
  Editor: KristoferDrozd
  Comment:
+   ← Revision 47 as of 2016-02-09 17:22:31 → ⇥
  Size: 4720
  Editor: DianeLambert
  Comment:
-Deletions are marked like this.
+Additions are marked like this.
 Line 1:
-= residuals.e =
+## page was renamed from residuals.e
= residuals =
-Line 3:
+Line 4:
-Residuals tests the landmarks to look for problems.  It will give you a report based on landmarks, pictures and maps.
+Compiled by KD

== Description ==

The routine residuals tests the landmarks to look for problems and gives reports about how well the landmarks were processed. In other words, residuals shows how well SPC performed. The reports that residuals outputs are listed in the Output section.
-Line 7:
+Line 12:
+ * [[NOMINALS]]
 * [[SUMFILES]]
-Line 8:
+Line 15:
- * [[PICTLIST.TXT]]
+ * [[PICTLIST.TXT]] - list of picture names generated by [[make_sumfiles]]
-Line 11:
+Line 18:
+=== Optional Files ===
 * [[PICTLISTS.TXT]] - user-generated list of picture names, if this file exists, it is used instead of [[PICTLIST.TXT]]
-Line 12:
+Line 22:
- enter plim (px,km,km)

 * The number of pixels an images needs to be "off" to throw the chevron (>>) flag
 * The rms error of a landmark 
 * Sets the "bin" size for the final histogram (basically how many images are in which side categories).
+ * enter 3 values (px,km,km)
  1. The value of the pixel residual limit. Images and maplets are flagged for a pixel residual limit greater than this.
  2. The value of the linear residual beyond which images and maplets are flagged.
  3. The max scale (km/px) that sets the scale for the histogram at the end of [[PICINFO.TXT]].
-Line 21:
+Line 29:
-.5, 1, 001
+, 0.01, 0.01
-Line 30:
+Line 38:
+ * [[check.txt]] - List of landmarks and overlaps for landmarks whose linear residual is greater than the user-specified limit.
 * [[EMPTY.TXT]] - [[lithos]] seed file for batch deletion of landmarks which are not contained in any pictures or limbs.
 * [[FLATLIST.TXT]] - List of flat maps (landmarks containing no topography).
 * [[LMKVECS.TXT]] - Landmark vectors for every landmark contained in [[LMRKLIST.TXT]].
 * [[MAPCHK.TXT]] - List of maplets whose difference between predicted and observed pixel/line locations in attached images is greater than a user-specified limit, or who have two or fewer overlaps attached to them.
 * [[MAPRES.TXT]] - Maplet resolution information for maplets contained in at least one picture or limb.
 * [[New_Limbs.in]] - [[lithos]] seed file to attach map to limbs for every landmark contained in [[LMRKLIST.TXT]].
 * [[no_update.txt]] - List of landmarks whose associated pictures are not listed by increasing mission time in the .LMK file.
 * [[PRUNE.TXT]] - List of landmarks for which the number of pictures in which they are contained exceeds the default limit of 500, or the user-specified limit, PRNLM, set within [[INIT_LITHOS.TXT]].
 * [[RANGES_SOLVED.TXT]] - Date and range information for each picture listed in [[PICTLISTS.TXT]] or [[PICTLIST.TXT]].
 * [[veto.txt]] - [[lithos]] seed file to detach from landmarks and limbs a map with a linear residual greater than the user-specified limit.
-Line 31:
+Line 50:
-=== Description ===
 * Compares predicted and observed pixel, line position of the landmarks in the images
 * Produces RSS of the diagonal convariance elements of each control point
  * Describes landmark position uncertainty
 * Flags landmarks with residuals over specified pixel value
+== Example of MAPINFO.TXT ==
{{attachment:residuals_controlpoint1.jpg||width=700}}
-Line 37:
+Line 53:
-=== Control Points ===
+== Example of a single landmark in RESIDUALS.TXT ==
{{attachment:residuals_imageresiduals.jpg||width=500}}
-Line 39:
+Line 56:
- * Inputs are wweights for S/C uncertainties from measurement uncertainties and from a-prior shape constraints
 * Control points solution results in a 3x3 output covariance for each control point. Diagonal elements are uncertainties along maplet coord. system axes, i.e. two horizontal directions & height. Typically SPC produces comparable horizontal & height uncertainties. Results are summarized as a scalar standard deviation per degree of freedom in meters
 * Residuals check automatically flags errors beyond certain input values in dump file for ease of inspection  
{{attachment:residuals_controlpoint1.jpg}}
+== What a Bad & Fixed Image Looks Like ==
-Line 44:
+Line 58:
-=== Image Residuals ===
+{{attachment:residuals_WhatWentWrong.jpg||width=700}}
-Line 46:
+Line 60:
- * 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 
{{attachment:residuals_imageresiduals.jpg}}
+== Qualitative Checks ==
The routine residuals is not the only way to check how well SPC performed. One can simply observe the maplets generated. The information below is what to look for when visually inspecting a maplet to judge its quality.
-Line 51:
+Line 63:
-=== Qualitative Checks ===
-Line 59:
+Line 70:
+== How Uncertainties are Introduced in the Data ==

 * Main types of uncertainties: 
  * S/C trajectory, camera pointing, image timestamp
   * Manifest themselves mostly in the projection of image template onto the surface.     Corrected by global geometry solution
  * Image noise, artifacts, smear, overall image quality
   * Manifest themselves in predicted image template brightness and in its fit to the extracted image brightness  
  * Photometric model and reflectance function models
   * Show up in slopes & heights integration
  * Poor choice of a-priori parameters & data weights
   * Evident at end of each estimation step; data won’t fit well

 * Often above contributions are correlated; individual contributors may not be separated until many processing steps have been taken