I am an undergraduate student starting work on my thesis using Cosi-Corr to assign ice velocities to Himalayan glaciers. First, I am attempting to find velocities of glaciers near Mt. Everest that already have published velocity data to ensure my methods are accurate. I am using two ASTER images that I've orthorectified and resampled using Cosi-Corr. I have correlated the two images and created a vector field using the data from that. I overlayed the vectors onto the images and would now like to extract velocity data from the vectors to assess the accuracy of my correlation so I can then apply this method to other glaciers. Is there anyway within Cosi-Corr or ENVI or maybe some outside way to figure out the amount of displacement that each vector in the field represents? I guess I am more or less confused on how to interpret the vector fields.

Thanks a bunch

Mark

I am doing a training project to calculate sand flux of a small area in Thar Desert, India. For this I am using Cartosat (2.5m) data. Due to the lack of ancillary file I could not  use COSI-Corr for orthorectification,but used another image processing software. However since the images were in the same coordinate system I used COSI-Corr to compute correlation (Freq correlator, 32 window size, step size 2). I then computed the Euclidean norm and subsequently calculated the sand flux. The problem I am facing is that the area has slow moving dunes and the sand flux values are coming out to be very high (about 5m per day), which is incorrect.
Could you please advise as to why it is coming like that? and what needs to be done? Is there some problem with the parameters taken during correlation or do slow moving dunes behave differently?

Please suggest
Thanks and regards,
Lydia

Excuse me for this heretical behaviour.

C. Boldo

I am trying to correlate two 2.5m spot images. The correlation step was ok (or so I thought) but I have the following error message when I try to replace SNR values < 0.8 by NaN :
"The following error was encountered: Array has too many elements. Please contact the supplier of the application."
After that, ENVI is shut down...

Any ideas?

Thank you in advance and have a nice day
Florian

For data of Cartosat 1 (2.5m), I have taken a variable window size of 64 to 32 pixels wide with step size 2; I have taken the Robustness Iteration= 2 and Mask Threshold (m)= 0.9. My query is:

Is Sinc resampling required for this data?why?
Instead of a variable window size, if I take a constant size of 32 pixels wide, would my results still be accurate?

Thanks in advance,
Lydia

I am using COSI-Corr in the context of glaciology and feature tracking.

I would like to know how the resulting rasters relate to the original input data. I have noticed that the results are smaller in size compared to the input images. When overlaying both input and output data, the output leaves a margin compared to the input (see example image: http://img337.imageshack.us/img337/9022 … sicorr.jpg). I believe this is due to the search window and the size of this margin is probably defined by the window size.

I would now need to know, how the pixel positions of the original image are represented by the result. In my example I used the frequency engine with an initial window size of 128x128 and a final size of 32x32. The XY step was 16 pixels. When I compare input and output I have a left margin of 127 px and a top margin of 128 px. However, when I calculated the correlation with the same setup and the same input images but without map information (non-referenced images), the resulting rasters were 1 px (16 original px) narrower. In this case the left margin was 143 px.

- How can I determine, which original pixels belong to each pixel in the result? In other words, is the magnitude of displacement calculated from exactly the 16x16 pixels 'underneath' the resulting pixel?
- How do the margins of 127 and 128 px come about?
- Why changes the output size when using georeferenced images?

Thanks a lot.

Firstly, i would like to point out that i am a new user of the CosiCorr. I am trying to determine the mass movements by cosicorr. However, i have got some questions. I would be too pleased if you could answer them.

I have got two ASTER (L1A *.hdf) images in two different periods. I want to determine the mass movements(about 5-10 meters) between two periods related to study area.

First of all, i wonder the answer of these questions.

I think we should use GCPS (5-6 GCPs in two images by Image to Image) in stead of orthorectification by ancillary file (about accuracy of 100 meters). We need also to a good DEM. Generally, you should suggest the shaded DEM for orthorectifying. We can have some difficulties to determine location of the tie points in shaded DEM.  On the other hand, in case of having the high resolution data (quickbird image, orthophoto or toposheet), can we use them for converting from tie points to GCPs? If we use them, Do we encounter any problems? During the orthorectification, if needed, which one of the optimized GCPS or GCPS should we use with DEM?

In correlation file or nlmfiltered file,  Data represents what (in Cursor/Location Value window),and what is its unit (metre or pixel)? For example, what does it mean in case of data:-12.125171.

Processes performing for my study

1. I applied the orthorectifications (by using ASTER GDEM) for images in two periods without GCPS.
2. I have the correlation file and it appears good.
3. I applied the filter the SNR Band of the correlation file (between 0.9 and 1)Discard / Replace images values
4. I defined subset area in the region without NaN values and determined  the h parametre for nlmfilter.
4. Although i applied nlmfilter for EW and NS bands, i have still some NaN values. How can we get rid of these NaN values?
5. Should i apply the destripe image for nlmfiltered image? if it should be, which one of files should we select for Automatic Orientation (in image destriping / undulation removal)

Thank you in advance.

This message to inform our members that Imagin'Labs Corporation (www.imaginlabs.com) has recently acquired the exclusive rights to exploits COSI-Corr for non-academic applications.

If you are interested in using COSI-Corr for non-academic applications, please contact Imagin'Labs Corporation (www.imaginlabs.com). Imagin'Labs proposes a wide range of services, including:

- Automatic image registration services,
- Image correlation services,
- Correlation of LiDAR or topographic data,
- Mutli-temporal and mutli-spectral analysis,
- COSI-Corr training, processing, and guidance.

For any question regarding commercial services, do not hesitate to contact info@imaginlabs.com

Cheers!
Sebastien

I am looking for any kind of support for using WorldView imagery in the NITF format. I am attempting to derive velocities for glaciers in Antarctica. I have a rough 150m DEM and no georeferenced imagery. It is not necessary that I tie the individual images to ground control points, only that the sequential images are tied together.

I guess I need to start with square one and am not sure what order to go in. There are about 40 images needed to cover the entire surface of the glacier. I can create swaths across the glacier, but then that takes 5-6 images. Where do I start with creating the ancillary files? Can I mosaic a swath first?

Any help or previous studies done using this software and worldview imagery is greatly appreciated!

Thank you!
~Brandon

I am trying to orthorectify and resample two SPOT 5 images (2006 and 2007). I begin with COSI-Corr.
I did the orthorectication/resampling step between my first image (2006) and a shaded DEM. It was ok. Then I selected 15 tie points between ortho2006 (the orthorectified one) and my second, raw image (2007). When I try to convert my set of tie points into GCPS, I got an error message if I use a DEM (I selected the real DEM, not the shaded one) :
"SYSTEM ERROR : TPOINTS_TO_GCPS--> Variable is undefined: TEXT."
If I do not select a DEM, no error message.
If I select 2006 image (not orthorectified) as my reference image, it works, but it says that my tie points are all outside the DEM...

An idea? Thank you in advance and have a nice day.

Florian

I have some ASTER images which their overlap area is very small. I cropped the raw images to the overlap area before tie point selection. It is because I think it´s not necessary for the whole image to be orthorectified. But I have a problem in GCP optimization step!! the GCPs can´t be optimized and I get NaN value as the coordinates.
Should I try to orthorectify the whole image extent to measure correlation in a small part of image?

any help would greatly appreciated

Thanks
Mohamad

Tony

its shows:

Trapped Error
SAT_GCPS_OPTIMIZATION:SAT_GCPS_OPTIMIZATION_CORE : Attempt to subscript <ptrHeapVar165> x_Pixel is out of range

I'm having this problem after resizing the reference image (Landsat) from 30m resolution to 15 m resolution to get the exact overlapping.

before resizing the reference and Raw images did not get exactly overlapped.

What i can do right now ? 

Sabber

ENVI Retrieve Data: An error has occurred during processing

Error: "Array has too many elements." The result may be invalid.

Any idea what may be causing this?

Pieter