ESA Mars Express VMC project (Version 3.1)


 

 

 

Introduction

 

The Visual Monitoring Camera (VMC) is a small camera on Mars Express. The astronomy group of the grammar school in Vaterstetten, Germany, was happy to get the opportunity to adopt a VMC-Operation. This operation took place on 22rd of March 2010 from 0:28 a.m. to 1:08 a.m. At this moment the satellite was situated halfway from apocenter of orbit to the planet. The aim was to process the raw - images supplied by ESA and then compare the images taken with the Visual-Monitoring-Camera (VMC) with our own telescope images and to generate a stereo image with the data of the VMC. So this was our work to do:

 

 

The students of the astronomy group are regularly meeting in order to observe celestial objects with the school’s telescope. On the web page of our school you will find astronomical pictures that have mostly been taken with a simple video camera by utilizing "Lucky Imaging". Some of these pictures are of breathtaking quality. There is a huge effort behind every self-made picture. The images of the Mars opposition in December 2007 taken with the Camera DBK 21 on Meade 30cm R-optics combined with Celestron 2x barlow are our best images of Mars so far. The seeing on 12/19/07 was good, 12/28/07 moderate, bad on all the other dates. Every picture in this set was extracted from a 5000-frame video. Proceeding from left to right the pictures cover almost the whole surface of Mars. In the lower left image you can divine Olympus Mons as a small bright spot in the upper left area of the picture. Even Arsia Mons is clearly detected by an experienced observer who knows its position (it is a very small bright spot 13mm below Olympus Mons slightly to the left).

 

Top of page

 

Processing Mars Express VMC Images.

 

Processing astronomical images is not easy. First you have to decide if you want to produce scientific images, or rather “pretty pictures”. In the first case you may not utilize many options that make pictures look nice by changing or loosing some of the original content. In the second case there is no such limit, goal is only a neat looking Picture.

We are of the opinion that making a picture “pretty” should also never change vital information, and change less vital information only to a very limited extent. We had this in mind when we processed the VMC pictures. This is a short description how we proceeded.

 

1. Converting data: All the .raw – files are copied to a folder containing the “vmc_flat.raw” flatfield and the “vmc2rgb.exe” file for converting .raw to .png - files. Then we draw every single .raw – file on “vmc2rgb.exe” with the mouse, the converted .png – file will show promptly. The VMC flatfield works fine, at the time being there is only one larger defect that is not corrected sufficiently. You should avoid choosing pictures showing this defect for your processing.

 

2. Selecting pictures: Camera position and attitude changes rapidly in a way that reducing noise by “Lucky Imaging” is no good idea with VMC data. We managed to use Lucky imaging successfully  for selected parts of the picture only, quite a demanding task. So we decided mainly to use single pictures for processing. The VMC data contain sequences of three or four pictures with different exposure times. We look for a nice one with high exposure, and another one with low exposure to fill out the overexposed and washed out area in the first picture.

Tipp: We heard about Software that will combine pictures with different exposure times to one high dynamic picture automatically. If you utilize such a software, use more than two pictures.

 

3. Histogram corrections: As the VMC pictures have pretty good levels, we make no correction. In other cases levels, contrast and intensity should be adjusted. Do not go to the limits, otherwise later operations may for example bring additional intensity and washed out areas. We do this adjustment at the beginning of  processing, because later on better visible structures will help to fit different images seamless together.

 

4. Sharpening pictures: We load the pictures by clicking “Datei/Bild laden” into the freeware program “Giotto 2.12” which has one of the best sharpening filters we know, the “Mexican Hat”. We click “Bearbeiten/ Schärfen und Filtern” and select the options “Quadrat, Bessel, Quadrat, Rauschfiltergrösse 3, Filtergrösse 5“ and try out the slider for “Filterwirkung“. “Filterwirkung” should be raised slowly, so that the picture gets sharp with not to much noise and artifacts. If the goal is a “pretty picture”, sharpening might be overestimated to a small extent, when a selective noise filter program is to be used in sequence. Processed files are saved in the .tif – format.

If  you do not intend to use a professional noise filter program, you should never sharpen too much in order to avoid nasty grain and artifacts. In this case continue with paragraph 6. If you are familiar with noise filtering, paragraph 5 is not a bad idea.

 

5. Noise reduction: We load the .tif – pictures to the program “Neat Image Pro+ 4.0”, which is suitable for selective noise filtering. To use this program is not an easy task. Users have to train with noisy images to find out. We can only give you a rough idea and tell how this works principally. First the program computes the “Device Noise Profile” of the image for the colors red, green and blue. you may change these profiles if necessary with nine sliders working for different intensities. Pushing up the left slider will for example intensify the processing in dark areas of the picture, note that you have to adjust the sliders for all three colors.

Then the user has to choose the “Noise Filter Settings”. you may adjust separately three ranges of frequencies: high, mid and low. For astronomical images the sliders for mid and low may be in the left position, as these frequencies are not of interest. We suggest using the Y Cr Cb working space. The sliders for “Noise Levels” and “Noise Reduction Amounts” must be carefully set, the results must be monitored in a rectangle drawn in the picture by mouse. Important is the correct adjustment of the Y – Sliders, because Y is mainly responsible for the sharpness of the picture. The Cr and Cb – sliders act on the color noise, which is less important for visual sharpness. If for example you visualize strong noise only in grey parts of the picture, the sliders in the “Noise Profile” concerning exactly this grey tone must be adjusted to a higher level. If you are sure to have the best result, choose “Output Image” and “Apply” to process the whole picture.

 

6. Building the picture: Remember we choose two pictures with different exposure times. Load both pictures to “Photoshop”. Resizing the pictures to 150% is convenient for processing and viewing the final product from a comfortable distance. Cut out the low exposed part you want to fill into the washed out area of the other picture with a 10 to 20 pixel feathered selection, draw it into the overexposed frame and adjust it properly to the correct position. If necessary put a layer with levels correction over the layer containing the cut out part, and adjust levels until there are no seams visible at all. If you have additional parts to fit to the picture (in picture no.2 we had to add the “horn” on both sides from two other frames), cut them out with feathered edges and fit them to the picture as described above. Concerning picture no.1 we could not resist to make a timid try on "Lucky Imaging": In a small region left and below of Arsia Mons we fitted a part from a third picture to find out what would happen. Indeed this region shows the best details in our results (see below).

 

7. Processing color: VMC pictures contain color which is not visible strongly. So adjust color saturation with Photoshop. If that will not do the trick, you may apply just a tiny little bit additional selective color to grey tones (neutrals), which is a method we do not like at all. We processed two different VMC pictures, no.1 had enough color, no.2 was processed a little bit to match the original color of no.1 picture. We do not encourage you to add a lot of color to a picture, that is not granted to be at least somewhat real.

Important note: If you are absolutely sure that all adjustments were made 100% correctly, and not the slightest hint of seams is visible, you should save the picture and keep it as a backup. Then merge all layers to one single layer to simplify coming operations.

 

8. Cleaning the background: Now we must clean up the background of  the picture. Put a selection circle exactly around the limb of the planet with a distance of about 3 pixels, and save this selection “circle” for further use. Then draw an elliptical selection and adopt it by using the selection transform tool thoroughly to the terminator of the planet with a distance of about 10 to 20 pixels. Be aware that later on this selection will be made feathered 10 to 15 pixels, which must adopt to the soft transition from light to darkness at the terminator. With our VMC pictures we were urged to use this selection to remove artifacts in the original picture which show as parallel lines close to the terminator. Save this selection “oval”. Now load selection “circle”, make it 3 pixels “feathered”, then load “oval” subtracting it from “circle”, and make it about 10 to 15 pixels “feathered”. Invert the selection, and redraw the background of the picture with a black or very dark brush. Clear the selection from the picture, and apply the brush with feathered tool tip to those parts you did not reach before, be aware not to come too close to the planet, there is no protection by a selection any more. For this operation it is a good idea to zoom the picture to a very large size and to use not to large, feathered brush tips for work close to the planet.

This sounds easy, but it is definitely not. It is very helpful if you establish a provisional layer for levels adjustment affecting all layers of the picture, an then pull the sliders to achieve a grey background with high contrast details. This will show clearly the effect of applying the brush to the background. If the planet is affected by the brush operation, cancel and readjust the selections in position and feather. The same is true if the brush will not come close enough to the planet. It might be convenient to draw over the limb or terminator with a red brush, to find out if the planet will be affected (don’t forget  to redo this operation). Save your work in between with different names, so a mistake will not ruin the previous work. If you have no practice, you will need al lot of patience to do the job.

 

9. The worst case: Avoid using the clone stamp or eraser to “correct” the picture, these are tools for the worst case only, but some times they are very useful. We find that a planet will only look good on the screen if there is sufficient dark background around. If necessary make the canvas larger and fill with background color. We are proud to claim the fact that the disk of Mars was not modified by our processing efforts, only a minor bit of color was added to the large picture, and the “horns” added were copied from images shortly before/after the main picture was taken. Display the picture filling the whole screen of your monitor, lean back and enjoy……

 

View our results:

 

     Upper line is image 1, lower line image 2.       Click the images to show original, click original to return.
   +  =
   Set your browser to full screen for viewing, often this works by pressing F11
   +

    =

   Image 3 was additionally  processed because of the interesting North Polar Cap.
 

    +

    =

   Image 4 was additionally  processed because of interesting haze in Valles Marineris.
 

08-283_04.27.15_VMC_Img_No_3

08-283_04.28.08_VMC_Img_No_4

08-283_04.29.01_VMC_Img_No_5

08-283_04.30.47_VMC_Img_No_7

 Blue and red

 stacked first

 =
   Images 5 and 6 taken only a few days apart were processed because of interesting haze and clouds.
 

10-317_00.33.20_VMC_Img_No_19

10-317_00.34.03_VMC_Img_No_20

10-317_00.34.46_VMC_Img_No_21

10-317_00.35.29_VMC_Img_No_22

 

10-331_00.15.34_VMC_Img_No_19

10-331_00.16.17_VMC_Img_No_20

10-331_00.17.00_VMC_Img_No_21

10-331_00.17.43_VMC_Img_No_22

   Image 7 compared with image 6 shows rapid change in Mars weather.
 

10-327_19.19.47_VMC_Img_No_19

10-327_19.20.29_VMC_Img_No_20

10-327_19.21.13_VMC_Img_No_21

10-327_19.21.56_VMC_Img_No_22

       
 

 

The picture of the North Polar Cap: On the way out to apocenter Mars Express VMC shot this wonderful picture of the North Polar Cap emerging from the dark terminator above. Composed of two RAW-frames it shows fine color shades and a lot of small craters not readily seen in the single frames. The lower left might show some dust, clouds or haze, the bottom right corner shows bright patches in the region of Moreux crater we suggest to be clouds.

When we saw the raw-material from 2010/08/08, we just had to give it a try. We decided to use two raw frames to reduce noise and to utilize the whole area of both pictures instead of selections only for better noise reduction. We resized to 150 percent and sharpened with a Mexican Hat filter. Then the two frames where ready for combining them.

The problem is, that due to rapid movement of Mars Express, the two Frames would not fit together. So we used Photoshop layer distortion to adapt one of the pictures exactly to the other, not an easy task. We accomplished sufficiently by "blinking" from one frame to the other, until blinking would not move the structures seen in the images. We then set the upper layer to 50 percent transmission and reduced the two layers to one. We then reduced high frequency noise with the software Neat Image, to do this without destroying to much sharpness is quit a demanding task.

Finally we used Photoshop to give the picture the "final touch" by carefully removing known artifacts of VMC, adjusting levels and color saturation and cutting the edges of the picture. When adjusting levels and color we used a mask for white with a tolerance of 100 and feathered edges of 30 pixels to protect the snowy areas of the Polar Cap in order to preserve the white color. This is  the link to ESA-presentation of our North Polar Cap work.

 

The picture of Valles Marineris: About 7500km out from Mars, Mars Express VMC shot a sequence of pictures showing interesting details. The whole system of Valles Marineris is drowned in a white haze. The large crater to the right seems to be Lowell crater, hiding its interesting inner ring-feature below haze. Originating from Arsia Mons a large cloud is visible that is getting lost in the dark of the terminator.

Viewing the raw-material from 2008/10/09, we had the idea the hazy patch seen at the terminator in the region of Arsia Mons could be some kind of reflection, but studying the other frames of the sequence showed that this rather is a true structure on Mars. Reaching into the dark of terminator it must be high above the ground, probably a large cloud originating near Arsia Mons. For processing it was decided to use four raw frames to reduce noise, taking into account that this would not deliver a perfect result due to the considerable movement of Mars Express during the 3.5 minutes slot the pictures were taken.

 

Using “vmc_flat.raw” flatfield and “vmc2rgb.exe” the pictures were converted from .raw to .png - files. The VMC flatfield works fine, at the time being there is only one larger defect that is not corrected sufficiently. After resizing to 150 percent, pictures 3/7 and 4/5 were stacked. After cutting out the overexposed part with a feathered selection the 4/5 result was stacked to the 3/7 result with 60% transparency. As expected the rapid movement of Mars Express became a problem, we could not find a perfect fit for the stacked frames. Having no special and expensive software for this purpose we accomplished sufficiently for a medium quality result only. The final picture was sharpened with a Mexican Hat filter, then high frequency noise was reduced with the software Neat Image, sacrificing some sharpness again.

Finally Photoshop was used to give the picture the "final touch" by carefully removing known artifacts of VMC, adjusting levels and color saturation and cleaning the dark background of the picture. When adjusting levels and color a mask for white with a tolerance of 60 and feathered edges of 3 pixels was used to emphasize the hazy area of Valles Marineris. This is  the link to ESA-presentation of our Valles Marineris work.

 

The three pictures "Hazy Mars": This very interesting VMC-material shows the almost complete north polar and surrounding area covered with clouds and haze. For all pictures the supplied darkframe “vmc_flat.raw” was used to extract png-files form the raw-material. Then we sharpened and stacked pictures No 19/21 and 20/22 in order to reduce noise.  After cutting out the well exposed part of the stack 19/21  this part was combined with the stack 20/22. The remaining noise was reduced further by utilizing Neat Image software. After cleaning the background, color saturation was adjusted. Finally for better viewing the result was resized to 125%, and some "fine tuning" done carefully. For detailed information on processing see our work above, and below concerning the best way of processing we worked out for the time being.

 

Picture 5: Only Korolev crater and Alba Mons could be found. The stunning details below Alba Mons could be high reaching and dense clouds with their shadows. These gigantic clouds extend roughly over 150km. Sand storms at some times may cover the whole planet with dust, but never before we had seen Mars with such an interesting cloudy and hazy area. Surface structure is clearly visible only in a small area around Alba Mons, even the north polar cap is not detected safely, although the pole is situated right on the terminator. The clouds show an interesting spiral structure, probably induced by coriolis force acting on air streaming out of a high pressure area on the northern hemisphere of a left spinning planet. Although clouds hide sharp surface-detail we decided to give it a try.

For picture  2010/11/13 we used:

 

Picture 6: This very interesting VMC-picture should be seen as supplement to previous image 2010/11/13. It shows the part of the polar region not visible in the 2010/11/13 image. Almost the complete north polar and surrounding area is covered with nicely structured clouds and haze. Identifying surface details is not easily accomplished. Only Acidalia Planitia is partly free of clouds, and Lomonosov crater  is detected easily. On a second look the large crater Lyot is seen full size inside a semicircle of clouds. Some other structures are easily detected by comparing with the Celestia image, but we do not know their names. Sand storms at some times may cover the whole planet with dust, but Mars with such a large interesting cloudy and hazy area is not seen every day. Surface structure is clearly visible only in small areas, even the north polar cap is not detected safely, although the pole is situated right below the terminator. The clouds show a stunning spiral shape, in a large stripe to the left very fine structure is visible. Although due to clouds there is little sharp detail we decided to give it a try.

For picture 2010/11/27 we used:

 

Picture 7: This picture combines two VMC-operations, the first 2010/11/23, the second took place only four days later 2010/11/27. Both operations meet a time with strong cloud and haze-activity on the northern part of Mars. Comparing these images my idea was to show the rapid change in cloud-structure. Due to the hidden surface it is not easy to identify landmarks, but I was abele to locate some prominent craters for better orientation comparing the two images.

Processing color from the original raw-frames by using the supplied flatfield, the atmospheric structures come out gray/white and not yellow/brown, so I assume they mainly are clouds and haze, not sandstorms. It is amazing how different these structures look, in some areas they look rather  smooth, in other areas they show very fine details. There also seems to be a difference between dawn and dusk, just compare the left (dusk) and right (dawn) terminator in the region of the “horn”. Also it seems, that surface conditions affect cloud structure above. Processing these pictures was not easy and took quite a while, but looking at the result I think time was not wasted. I do like this picture. Only operation 2010/11/23 is new to picture 7:

For picture 2010/11/23 we used:

 

 

Processing is sometimes not easy, and quite a bit of practice is helpful. Everybody may use his favorite software and try out what he can do with it. We use Photoshop, Giotto “Mexican Hat” for sharpening and Neat Image for noise reduction. Here I want to recall the single steps I used processing the actual picture, this seems to be the best way of processing we worked out for the time being.

Remarks: We do not use the library-png because we want to see the color of the raw-material and adjust it so that for example white clouds stay something near white. We sharpen every single picture, doing this a bit of noise is no problem because later two frames are stacked, and a professional noise filter is applied. Artifacts must be cleaned from each picture before stacking, choosing pictures with artifacts not in the same place will preserves some original information for all parts. We align two frames in order to reduce noise, for this usually one image must be rotated a bit. We stack separately for high and low exposure. Fitting the high exposure section to the low exposure picture is done by feathered selection and must be tried out very carefully.

Cleaning the background is done by a 2px feathered circle-selection very close to the rim of Mars, then a 15-20px feathered oval selection is used at the terminator. This must be done carefully not to alter the original picture more than necessary. If adjustment of color saturation results in a bad color noise, this may be reduced by applying a noise filter to the color information only (definitely not affecting the luminance). For this we separate luminance and color by Photoshop. At the end some “fine-tuning” may be favorable (for example feathered selection on a cloud patch to enhance white color and so on). We do this carefully and moderately in order to preserve the original content of the picture.

 

Top of page

 

VMC camera versus backyard telescope

 

We tried to take pictures of Mars with our backyard SC-optic featuring an aperture of 30cm. These low cost optics are not of superior quality, but after several tries we finally selected a pretty good one in 2007, and could make nice pictures of Mars. A single frame taken would due to atmospheric seeing, atmospheric dispersion and minor issues with optics quality of course not reach the theoretical limit of resolution, but utilizing “Lucky Imaging” (processing 5000 frames for one picture) we managed to compensate these disadvantages (the principal approach is shown here). So keep in mind for the  following discussion, that the displayed pictures of Mars opposition 2007 shown in the introduction come close to the theoretical limit of 0,4” resolution of our telescope. Now let us compare VMC and SC resolution:

Image scale in km/pixel: It is easy to find the image scale S in km/pixel by  S = (x * d)/f  where d is the distance from the object, f is the focal length of the optic and x is the pixelsize of the camera chip. Corresponding to our image no.1 we assume an altitude of 9947km above Mars for the VMC camera.  The VMC focal length is 12,3mm, pixelsize is 14 micrometers. The data for our telescope image: distance to Mars 88200000km, pixelsize of DBK 21 camera is 5,6 micrometer, and 6m focal length of our Telescope. So we find:

                   S1 = (0,014mm * 9947km)/12,3mm = 11,3 km/pixel  for the VMC

                   S2 = (0,0056mm * 88200000km)/6000mm = 82,3 km/pixel  for the telescope

Putting things together: This shows that the resolution of the VMC image should be a factor of  82,3/11,3 = 7,3  better than our telescope image. The Bayer mask for color imaging will degrade image resolution by a mean factor of 1,3 (look at the edge of a razor blade imaged with the same camera with/without Bayer mask). This will not leed to mistakes calculating relative resolution, because VMC and DBK 21 both have a Bayer mask. Remembering the statement above that in the telescope picture seeing etc. is compensated by “Lucky Imaging”, and assuming that movement during exposure time may be neglected or will have similar effect in both cases, there is no other strong influence involved, the factor 7,3 is the final result for good observation conditions on earth.

Important note: The above computed image scale will only then equal resolution of the system, if the airy disk of the camera optic is not to large compared to the pixelsize of the sensor.  We checked this thoroughly and found no problem. Doing this we utilized a nice and easy to use formula  published on our homepage which is F = 2,2x to compute the minimum aperture ratio F for a chip with x = 18,2 micrometers effective pixelsize (VMC pixelsize 14 micrometers corrected by the Bayer degradation of 1,3). So for a given aperture of 2,46mm VMC should be operated at a focal ratio of F = 2,2 * 18,2 = 40 for best resolution in planetary imaging. That is eight times larger than the actual F = 5. For the aperture of 2,46mm the focal length should be 8 * 12,3mm = 98,4mm to utilize theoretical resolution for planetary imaging. This shows clearly, that VMC was originally to be used for monitoring and not for planetary imaging.

 

What the pictures tell: Be aware of the fact that all following diameters correspond to viewing the images on our monitor. Choosing other screens will change all measured diameters, but will not change the results. It is not easy to measure the finest structures in our telescope images, but the size definitely is close to 1,6mm for a 50mm disk diameter of Mars (which corresponds to a resolution of 0,5”). We find about 1,0mm for finest details in the 200mm disk of the WMC image we processed. We must take into account that this 200mm will be somewhat smaller than the full disk of Mars, because we are to close to the planet to view the full diameter. 210mm will be a good estimate for the full disk. Reduced to a 50mm disk finest details in the VMC image will be 1mm * 50/210 = 0,238mm, this is compared to the size of 1,6mm from the telescope image a factor 1,6/0,238 = 6,7 in favor of VMC, which for a rough estimate matches fine to the factor 7,3 we found above. Our image no.2 was taken by VMC closer to Mars, due to about half the altitude resolution of the picture should be up to 14 times better than our telescope image.

 

 

Finding that VMC in apocenter is about seven times better than our telescope, we recognized that Hubble WFPC2 also should be that much better: The 2,40m Hubble Telescope should be eight times better than or 30cm SC (for small optics resolution is in the first place a function of the aperture only). Assuming that Hubble WFPC2 planetary camera operated at F = 28,3 in visible light is not matched absolutely perfect to the 15 micrometer pixelsize of the CCD sensor, this will compensate for degradation by Bayer mask in our system. So a Hubble image of Mars taken near opposition 2007 should be comparable to our VMC picture in apocenter. The proof might come from the above picture, where we adjusted color in both Mars-frames for a better match.

When Mars Express leaves apocenter and approaches the Planet, resolution of WMC images will increase so much that even the powerful Hubble Telescope has no chance, this shows drastically the importance of missions like Mars Express, if  you want to find out, you must go there.....

 

Top of page

 

Stereo Images

 

Calculating the best resolution of VMC pictures at apocenter observation point gave a result of only 11km/pixel for the central part of the picture, the negative impact of Bayer mask leading to additional degrading. This led to the conclusion the stereopictural portrayal of Mars relief would hardly be possible. The received data and the described attempts of increasing the photographs’ quality proved this assumption true. For this reason the three-dimensional projection of the spherical shape of the planet seemed to be much more reasonable. More profound considerations showed, that because of the great distance of the satellite from the surface of Mars no three-dimensional images would be possible. It is the same reason why the moon looks like a two-dimensional disc when observed from earth. But it's absolutely accepted in telescope observations to look at planets or the moon for better viewing through a binocular adapter as seen on the right telescope in the picture. This enables the observation of an object by two different optical paths with both eyes at the same time. The observer gets a three-dimensional impression even though the object is viewed through the same telescope aperture. So to produce the stereo image we simply used two identical images.

 

In the above picture the correlation of the two left pictures may be caused by squinting. Your brain will, from the information gathered by evaluating the switched position - images, produce a 3D–picture. Please adjust distance and position of your head carefully, until there is no stress squinting at the pictures. With some training you will notice finest details in the image not seen readily by "normal" viewing. Just give it a try, not everybody will succeed. You also have the possibility to use glasses parted in red-cyan to get a 3-dimensional impression of the right stereo image made by using StereoPhotoMaker, but the squinting method is better.

If the trick works, you will recognize why astronomers like to use binocular adapters to view Moon and planets. On a telescope you don`t have to squint, and you will notice that your brain will enlarge the image even more, than squinting makes it smaller. Viewing the Moon with such a telescope is one  of the most impressive things you can do in astronomy. This is  the link to ESA-presentation of our adopted VMC-opertation. The ESA presentation of our additional work is found here: Three Views of Mars  stunning  North Polar Cap interesting  Haze in Valles Marineris.

 

 

 

Conclusion

 

The Humboldt-Gymnasium-Vaterstetten’s astronomy group finally wishes to thank the ESA and Mars Express Flight Control Team. for VMC observation time and given  possibility to change perspective from the school telescope’s point of view to a space satellite at a very close range to the investigated object, to work with the received data and publish its results and methods to a community that may take advantage of it. Special thanks go to Peter Wellmann, who helped us a lot with his great experience in image processing.

 

 

Supplement

 

 We are very delighted by the extraordinary acknowledgment of our work by ESA. ESA wrote:

 

"Speaking on behalf of the entire Mars Express Flight Control Team, I am very impressed with the work done by the teachers and students at the Humboldt Gymnasium. Their work, analysis and results prove the value both educational and scientific of even 'low-tech' images delivered from deep space. Congratulations on a project well done and we wish you continued success in your studies."

--- ESA's Michel Denis, Mars Express Spacecraft Operations Manager, ESA/ESOC


All of us here at the Mars Webcam blog were tremendously impressed with the work done by the students. The goal was to analyse VMC images and determine how these compare in resolution to images obtained from the ground and, interestingly, from the joint ESA-NASA Hubble Space Telescope.
The student team was able to demonstrate that the VMC camera, viewing Mars from 10 000 km, provides images having similar resolution to those provided by the Hubble telescope viewing Mars at 88 million km. They also created an excellent stereo image of Mars.

--- Congratulations and thanks for an excellent report!

 

Top of page