Working with Mars Viking Orbiter Data

About the Viking Mission to Mars

Processing Viking Orbiter VIS Data

• Level 0: Data Ingestion
  Acquire Viking VIS data from one of numerous sources, import it into ISIS, and initialize it with SPICE information.

• Level 1: Noise Removal and Radiometric Calibration
  Remove the salt-and-pepper noise, reseau marks, and missing track noise in the data, then radiometrically calibrate the image data so the DNs represent reflectance (ranging from 0 to 1).

• Level 2 - Projection
  Geodetic corrections are performed and the images projected to a map projection.

• Level 3 – Photometric Correction and Enhancement
  The effect of sun angle on the image is corrected, and the images are tone matched.

• Level 4 – Building a Mosaic
  A seamless mosaic is created.

Level 0 Processing - Data Ingestion

This is the starting point for the production of a Viking mosaic. The steps within the level zero processing provide the gateway into ISIS processing. Running the following applications will ingest the Engineering Data Record (EDR) and place necessary information into the labels of the image. Viking Obiter information has been around for quite awhile so problems with the dataset have been discovered and accounted for, allowing this process to run smoothly.

Acquiring Viking Orbiter Data

• NASA PDS (Home)

  • Image Atlas Search - Viking
  • Data Holdings - Viking
  • Online Data Volumes - Viking

• WUSTL Mars Orbital Data Explorer

• ASU Viking Orbiter Image Map

• ASU JMARS

Widen the search area to find shifted images

When searching for Viking orbiter data that covers your area of interest, it can be useful to widen you search area by as much as five degrees to each value in your latitude and longitude search values. The camera pointing for Viking images may be off by as much as a half frame so expanding your search will help insure that you get all the images that were acquired for that area.

Data Acquisition Tool: PDS Image Atlas

Search

The PDS Image Atlas can search for Viking Orbiter Imagery. Here are some relevant filters:

Parameter	Notes
Mission	'Viking Orbiter' to focus on images from this mission.
Target	'Mars' to filter out Viking images of other planetary bodies.
Product Type	'edr' The type of Viking data to search for.
Lat/Lon Bounding Box	Enter min and max latitude and longitude values that box in your area of interest.

Browsing Manually

The PDS Data Holdings offer Viking data archives by data type.

The PDS Online Data Volumes offer an archived copy of the CDs and tapes originally used to catalog and distribute Viking data.

• See Also: Planetary Image Atlas Help

Data Acquisition Tool: JMARS

JMARS is a Geographic Information System (GIS) tool. It can be used to evaluate Viking images that cover an area of interest before the images are downloaded. The tool displays Viking footprints and a variety of other Mars data. JMARS can:

• query the database of Viking images,
• select browse images to display on-screen and download via the web,
• generate a list of Viking images (great for creating scripts), and
• save the displayed map as an image.

Using JMARS to Display Viking Footprints

1. Once JMARS is installed, launch it and log in.
2. In the Main tab of the Layer Manager, hit the Add new layer button, which opens a menu.
3. Select Viking from the Stamp menu.
4. In the Add Viking stamp layer window, just hit the Okay button, leaving all the fields blank.
5. A new tab named Viking stamps will appear in the Layer Manager. When it's done loading the footprints, its drawing status indicator will turn from red to green, the footprints will be displayed on the map and the images shown on the map are list in the Layer Manager.
6. Right click on a listing in the image list and on the footprints displayed on the screen - there's lots of options for working with and accessing information and data. For example, Render and Load Selected Stamps will download and display the Viking images for the selected footprints on the map, giving you the ability to preview the data. Web Browse will launch an image's web page in your browser so you can access the information and data.

Other options in JMARS to help you search for Viking data include the tools for narrowing your search and modifying your display in the Settings, Query, and Render tabs in the Layer Manager, the various tools in the main menus, and adding other data layers to the display.

Screenshot of JMARS displaying Viking 2 footprints:
The Viking footprints (called stamps in JMARS) are displayed as blue polygons on the map. Several footprints are shown selected in the Layers Manager Viking 2 Stamps list and highlighted yellow on the map. Several Viking images are displayed the map, filling in those footprints with a preview of the actual image data. Image names for the selected footprints have been copied from JMARS to our text editor. In this case, we selected Viking 2 for our new layer.

• See JMARS (ASU)
  Free, registration required for full functionality

Importing Viking Orbiter Data

Viking VIS data are distributed in Standard Data Products formatted files, which have an .imq extension. These are compressed PDS format images, and are decompressed before ingesting into ISIS.

Ingest VIS Images into ISIS with vik2isis

vik2isis FROM=Viking_input.imq To=Viking_image.cub

Decompression for .imq images is included in vik2isis, no need to do anything extra.

Viking VIS Problem Data - Clouds and Dust

Some problems with the Viking data set are due to environmental conditions, and the easiest way to spot them is to visually inspect the images. Clouds and airborne dust degrade the quality of the image.

The decision whether to use image is a judgment based on how much information will be gained vs. how much the image will degrade the final product.

• 

  VIS Image with Clouds

• 

  VIS Image with Dust

Adding SPICE

About SPICE

ISIS geometrically and photometrically characterizes pixels in raw planetary instrument images. Information such as latitude, longitude, phase angle, incidence angle, emission angle, local solar time, sun azimuth, and a many other pixel characteristics can be computed.

For ISIS to work, the SPICE kernels must first be determined for the particular raw instrument image. These kernels maintain the spacecraft position and orientation over time as well as the target position and instrument specific operating modes.

Add SPICE information to your cub with spiceinit

To add SPICE information to your cube, run spiceinit on the image so other ISIS apps (i.e, cam2map, campt, qview) will have the info they need. Only the FROM parameter is needed:

spiceinit FROM=my.cub

Level 1 Processing - Noise Removal and Radiometric Calibration

To create a Level 1 Viking image, clean up noise and imaging artifacts, then radiometrically correct the data to get an image representing the reflectance of the surface.

Noise/Artifact Removal

• 
  Remove white 'salt noise' speckles

• 
  Remove reseaus

• 
  Fill in missing tracks

• 
  Remove black 'pepper noise' speckles

Removing Salt and Pepper Noise with viknosalt and viknopepper

The black and white speckle in Viking images is a result of interference during the transfer of information from the spacecraft. This speckle is called salt and pepper noise because it has the appearance of grains of salt and pepper sprinkled across the image.

Removing Salt Noise

viknosalt runs noisefilter five times to identify white noise within the image. The identified pixels are set to a value of null. The final step in viknosalt is a low pass filter that replaces the null pixels with a valid value.

viknosalt command:

viknosalt FROM=Viking_image.cub TO=Viking_nosalt.cub

• 
  Before removing salt noise

• 
  After removing salt noise

• 
  Close-up before salt noise removal

• 
  Close-up after salt noise removal

Removing Pepper Noise

See the viknopepper ISIS application docs for info on removing pepper noise. It works similarly to viknosalt.

Removing Reseaus

Reseau marks on Viking images are small dots that form a grid of points across the image. These marks are created by design -- reseaus are etched in a pattern over the lens of the camera, and the marks the reseaus make on the image will allow us to refine the image in later processing steps. The images below show reseau marks on a Vidicon tube similar to the ones used for the Viking VIS cameras, and an enlargement of a reseaus mark in a Viking image.

• 
  Reseaus on a Vidicon tube

• 
  Close-up of reseau mark on a Viking image

The next step is to find and remove reseaus. The reseau locations found in this step are used to correct for the optical distortion. The vidicon cameras used by the Viking spacecraft have electronic distortions similar in pattern to the optical distortion in a film camera. These electronic distortions are introduced because the scanning pattern of the electron beam used to read out the charge-stored image vidicon is more "barrel-shaped" than rectangular. Interactions between the charge on the photo-cathode that represent the image itself and the electron beam produce additional complex high-order distortions.

Reseaus removal is accomplish by using two applications: findrx to find the reseaus, then remrx to remove them.

Finding Reseaus with findrx

findrx will read in a cube and refine the position of the reseau points based on information about where the reseaus should be and comparing those areas of the image against a Viking reseau pattern to locate the actual reseau mark in the image. The image labels are then modified to reflect the new sub-pixel accuracy.

findrx command:

findrx  FROM=Viking_nosalt.cub

Removing Reseaus with remrx

remrx removes reseaus from a Viking image. When you select a value for the parameters sdim (sample dimension) and ldim (line dimension), you want to choose values that are large enough to remove the reseaus but, the value should not be larger then required or you will remove valid data. For most Viking images we have found that of sdim= 9 and ldim= 9 works well. While the reseaus are visibly removed from the images, the reseau information is retained in the labels for later processing stages.

remrx command:

remrx FROM=Viking_nosalt.cub TO=Viking_norx.cub sdim=9 ldim=9

• 
  Input image (Viking image with salt noise removed)

• 
  Reseaus removed using sdim=9 and ldim=9

Comparisons of remrx with different ldim/sdim parameters

Remove Missing Track Noise with vikfixtrx

The missing track noise is caused by interference from the spacecraft electronics. Track noise appears as a regular pattern of NULL values across the image in increments of seven pixels. This noise occurs because of the way Viking images were transmitted to Earth. On the spacecraft, an image is broken into seven tracks, each track contains every seventh value along a scan line. If errors occurred while a track was transmitted to Earth, then the track contains incorrect data values. Sometimes, several tracks are missing.

vikfixtrx checks for this type of noise in each of the seven tracks of a Viking image cube. If the threshold of invalid pixels in a given track is met or exceeded, then the track is considered bad and all pixels are replaced by interpolating valid values from either side of the bad pixels.

vikfixtrx command:

vikfixtrx FROM=Viking_norx.cub TO=Viking_notrx.cub

• Input image (Viking image with reseaus removed)

• 
  Output image (null tracks removed)

Close-up Input image

Close-up Output image (null tracks removed)

• See Also: General Overview of Noise & Artifacts

Radiometric Calibration

• 
  Radiometrically calibrate the image

Radiometric Calibration of Viking VIS Data

vikcal performs radiometric corrections to planetary images acquired by the Viking orbiter cameras. It performs a radiometric correction in two steps:

1. Correct the varying response of the vidicon across the field of view of the camera. Multiplicative and additive correction coefficients, as a function of line and sample position, are applied to an image array to produce the results of an 'ideal' camera.
2. Convert the image data to reflectance values, where reflectance (a value between 0 and 1) is the ratio of the observed radiance and the radiance of a white screen, normal to the incident rays of the Sun.

vikcal command:

vikcal FROM=Viking_notrx.cub TO=Viking_cal.cub

The output image of vikcal will be a 32-bit (floating-point) cube, where the pixel values represent reflectance.

• See Also: General Overview of Radiometric Calibration

Level 2 Processing - Geometry

Producing a mosaic requires geometric processing on the individual images that make up the desired mosaic. The individual images are geometrically transformed from spacecraft camera orientation to a common map coordinate system. ISIS has generic software applications that are applied to all supported mission data. Based on the information in the image labels that was added in our earlier steps, the software recognizes the instrument and performs accordingly.

• See Also: Map Projecting Images

Level 3 Processing - Photometric Correction

Currently, ISIS photometric correction capabilities are under development. When the applications are released, we will develop a page here in this lesson providing you with examples and tips for using ISIS photometric correction tools.

Photometric normalization is applied to all images that make up a mosaic in order to adjust and balance the brightness levels among the images that were acquired under the different lighting conditions.

Generally, radiometrically calibrated spacecraft images measure the brightness of a scene under specific angles of illumination, emission, and phase. For a planetary body that doesn't have a significant atmosphere, this brightness is controlled by two basic classes of information: the intrinsic properties of the surface materials , (including composition, grain size, roughness, and porosity), and local topography of the surface.

Level 4 Processing - Mosaicking

Currently, ISIS photometric correction capabilities are under development. When the applications are released, we will finish developing this lesson and provide you with tips for using ISIS to create your final, seamless mosaic using mapmos and tone matching procedures and applications.

The final steps in our processing will produce a seamless mosaic of all the images in our region of interest. In spite of best efforts at radiometric calibration and photometric normalization, small residual discrepancies in image brightnesses are likely to remain. These brightness differences appear as seams in a mosaic. There are a couple of methods that will minimize the seams resulting in an improved aesthetic result for a final mosaic. The accuracy and quality of the radiometric calibration and photometric normalization effects how well the seams can be minimized.

Exporting ISIS Data

• See: Exporting ISIS Data