PixInsight Forum

PixInsight => Image Processing Challenges => Topic started by: Carlos Milovic on 2014 April 16 09:07:07

Title: Deconvolution Challenge
Post by: Carlos Milovic on 2014 April 16 09:07:07
Today we start a new challenge, to test our new deconvolution tool against the old ones. So, we ask you to give your best shot to the test images that we'll post here. The best results are going to be included here. Please try to use all the algorithms avalaible right now (Constrained Least Squares or Wiener filters in RestorationFilter and Regularized versions of Richardson-Lucy and Van Cittern in Deconvolution, specially).

Test 1: Bigradient image - without noise
This is a simple test, without noise to complicate things. Do your best to get sharp edges and as less ringing artifacts as you can.
Image: http://pteam.pixinsight.com/decchall/bigradient_conv.tif
PSF: http://pteam.pixinsight.com/decchall/psf.tif


Test 2: Bigradient image - with gaussian noise
To make things more difficult, let's add a bit of noise to the image. In this case, 10% of gaussian noise. Please, do not use noise reduction algorithm here.
Image: http://pteam.pixinsight.com/decchall/bigradient_convnoise.tif
PSF: http://pteam.pixinsight.com/decchall/psf.tif


Good luck, and thanks for giving it a try!!!
Title: Re: Deconvolution Challenge
Post by: Alejandro Tombolini on 2014 April 16 16:13:03
Hi Carlos, I think that it is almost impossible with the actual tools to recover the original information of the image  >:D

Without noise I have applied a convolution to reduce the lines generated for applying Van Cittert in Deconvolution without regularization.
With noise I also used Van Cittert but this time with regularization.

Waiting for the new tool!!!  :)

Saludos, Alejandro.
Title: Re: Deconvolution Challenge
Post by: Carlos Milovic on 2014 April 17 08:39:27
Thank you for your results, Alejandro! It is not much surprise that Van Cittern performed better here than other algorithms (RRL and CLS/W). If we remember correctly, Van Cittern outperforms them in lunar images, which are quite similar to this test: strong edges and smooth, quite constant features.

The new deconvolution tool, TGVRestoration is right now on the optimization stage. Meanwhile, we are trying to know how the parameters behave and how it stands against our previous tool. Just as a token of what is comming, here is the result of the first challenge (without noise). Please, keep the results coming. I'll post the result of the second challenge in a couple of days.

(http://pteam.pixinsight.com/decchall/bigradient_t1_tgvr.png)

As you can see, there are only small artifacts at the corners of the square.
Title: Re: Deconvolution Challenge
Post by: Philippe B. on 2014 April 17 08:59:07
It looks promising !!!!!  :tongue:


Thanks for these exciting new features >:D

Please optimize quickly ! Cannot wait 3 months
:laugh:
Title: Re: Deconvolution Challenge
Post by: Alejandro Tombolini on 2014 April 17 10:11:58
Hi Carlos, the only way of not to say all I should to represent how excited I am with this new tool is with a Spock-like answer. "Fascinating"

Saludos, Alejandro.
Title: Re: Deconvolution Challenge
Post by: jeffweiss9 on 2014 April 17 13:09:10
Carlos-
 As someone who has found the deconvolution process in PI to be the perhaps the most difficult to master, I can hope that the new deconvolution will be a little more bullet-proof for the more pedestrian user.  I've never been able to get consistent results with it except for a few times where I hit the lottery on my parameter and mask settings. Most of the time I am unsuccessful and, if I deconvolve at all, have to use AIP4WIN Lucy-Richardson which, if nothing else, operates consistently and produces results for any image I give it. This is the last processing step I've never been able to do (consistently) in PI.
-Jeff
Title: Re: Deconvolution Challenge
Post by: Enzo De Bernardini on 2014 April 18 17:49:28
My try (bigradient_conv)

Impressive results Carlos!.

Regards,

Enzo.
Title: Re: Deconvolution Challenge
Post by: Carlos Milovic on 2014 April 19 08:17:41
Here is the result of the second challenge, with TGVRestoration:

(http://pteam.pixinsight.com/decchall/bigradient_t2_tgvr.png)

There are more artifacts left than in the first challenge, but IMO this new tool outperforms quite well the other algorithms.


Jeff, I had the same feeling about deconvolution in PI. It was really hard, specially setting the deringing parameters. This new deconvolution is not artifact or ringing free, but so far yields much better results. If you are familiar with TGVDenoise, then this tool will be straigtforward to use. The strenght parameter behaves a bit different, but at the end there are the same relevant 3 parameters (strength, edge protection and smoothness). I've included also the same deringing algorithm as in the previous deconvolution, since we cannot quarantee that the PSF model is accurate, so ringing may happen anyway... but, so far in my tests, I have not used this option. So, what you have seeing here is just fine tunning the same 3 parameters as in TGVDenoise (given that you know/have the PSF).

Title: Re: Deconvolution Challenge
Post by: astroedo on 2014 April 21 09:52:28
I CAN'T BELIEVE IT!!!!!!!  :surprised: :surprised: :surprised: :surprised: :surprised: :surprised: :surprised: :surprised: :surprised:

I can't wait to see it in action on planetary and deep sky images!

looking forward!

Thank you guys!
Title: Re: Deconvolution Challenge
Post by: Juan Conejero on 2014 April 21 16:45:23
...I've never been able to get consistent results with it except for a few times where I hit the lottery on my parameter and mask settings.

As the guy who wrote the Deconvolution tool eight years ago, I can feel involved with the implications of this assertion. Can you upload one of those images where you are getting inconsistent results?

Quote
if I deconvolve at all, have to use AIP4WIN Lucy-Richardson which, if nothing else, operates consistently and produces results for any image I give it.

Please realize that we are not amateurs, so this is not a hobby for us. Software development is our profession, and PixInsight is a professional software development project. Competing applications and their developers are not our "friends" or "buddies". Since we have to pay invoices and salaries we have to sell licenses, so there are no jokes here.

So naming other applications is not the most productive way to express an adverse opinion about our software on this forum. Asking specific questions, uploading data for evaluation, reporting bugs, requesting new features or just criticizing us constructively, are much more efficient options.
Title: Re: Deconvolution Challenge
Post by: jeffweiss9 on 2014 April 21 19:37:23
Juan-
   I just uploaded the image I (last) was unable to deconvolve properly (m8182LStk35cr_DBE.fit which is the one; but also uploaded a version prior to DBE), in jweiss directory in forum shared files. It is the result of 35 8-minute Luminosity subs of M81/M82 with IFN and SN2014J through BatchPreprocessing and separate ImageIntegration.   I didn't mean to offend you, although obviously my quoted statements had that effect.  I would love to learn a methodology that a time- and patience-limited amateur like myself can handle.  I am trying to achieve my goal of ridding my workflow of its 1 or 2 residual non-PI steps since I do believe PI has the best algorithms, in general, that are out there, once I'm able to master them.
   Respectfully,
Jeff
Title: Re: Deconvolution Challenge
Post by: Andres.Pozo on 2014 April 22 01:59:46
I have a suggestion for the new convolution process: In many of my images the right value for the deringing parameter is less than 0.01. However the slider has a precision of 0.01 units. So for selecting a smaller value I have to write it on the textbox. I think that you should set a smaller range in the slider and increase the precision, i.e. min:0 max:0.2 inc:0.001. That is only 200 steps.

Also, the default value is 0.1. This is way too much for most linear images. I have seen users complaining about that deringing destroys their images, and I think it is because of the default parameter. I think it would be better to have as default a value that does little than a value that wrecks the image.
Title: Re: Deconvolution Challenge
Post by: Juan Conejero on 2014 April 24 05:44:31
Jeff,

Thank you for uploading your data. This is a brief tutorial with your image. I have tried to follow a systematic step-by-step procedure to apply our deconvolution tools in a way that can yield consistent results for most deep-sky images.


Step 1: PSF Estimation

Our deconvolution tools (including the incoming TGVRestoration tool) don't implement blind deconvolution algorithms. This means that an accurate estimate of the PSF of the image is crucial to perform a meaningful deconvolution. Of course, I don't need to say that these processes, PSF measurement and deconvolution, only make sense with linear data.

DynamicPSF is the tool of choice for PSF estimation in PixInsight:

(http://forum-images.pixinsight.com/20140424/Deconvolution/01-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/01.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/01.jpg)

I have selected a total of 52 stars. As you know, we normally don't need hundreds of stars to get a good PSF estimate. Just a few tens of carefully chosen stars (unsaturated, neither too bright nor too dim stars) are sufficient. From the 52 stars I have selected the best 40, after sorting the list by mean absolute difference.

(http://forum-images.pixinsight.com/20140424/Deconvolution/02-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/02.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/02.jpg)

Then I have obtained the PSF estimate as a new image with DynamicPSF's Export synthetic PSF option.

(http://forum-images.pixinsight.com/20140424/Deconvolution/03-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/03.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/03.jpg)


Step 2: Linear Mask

Deconvolution must only be applied to high-signal areas of the image. This clearly excludes the sky background and dim structures such as IFNs in this case. These marginal data components can be protected with a suitable mask. Normally this step is not necessary with our Deconvolution tool, since regularized algorithms are already very efficient at preserving nonsignificant structures, preventing noise amplification on these regions. However, I prefer to include a mask generation task in this case to provide you with a more complete example.

Typically, background protection masks are generated by applying nonlinear transformations, such as a histogram transformation with a non-neutral midtones balance, or a gamma stretch. This generates a nonlinear stretched version of the image that tends to protect dark regions. Actually, this is a conceptual error: If we want to protect image structures as a function of the noise-to-signal ratio, we need a mask where pixel values are a function of the signal, but with a nonlinear mask generated as described above, this cannot be guaranteed in general.

A linear mask, that is a mask where linearity of the original data is preserved, is an accurate and easy-to-build protection mask for this purpose. Linearity means that only linear operations must be applied to a duplicate of the image. We begin with a straight multiplication:

(http://forum-images.pixinsight.com/20140424/Deconvolution/04-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/04.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/04.jpg)

The constant 60 acts like an amplification factor. The brightest areas of the image become completely saturated (white), which means that they will be fuly deconvolved. The next step is a linear histogram clip at the shadows:

(http://forum-images.pixinsight.com/20140424/Deconvolution/05-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/05.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/05.jpg)

and a convolution with a small Gaussian filter. This makes the mask more robust to local noise variations. The smoothing applied must not be too strong, or the mask will become inaccurate (for example, dim stars can become wrongly protected if the mask is too smooth).

(http://forum-images.pixinsight.com/20140424/Deconvolution/06-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/06.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/06.jpg)

Finally, the mask has to be activated for the target image:

(http://forum-images.pixinsight.com/20140424/Deconvolution/07-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/07.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/07.jpg)

As I have said, a protection mask is normally not necessary with our Deconvolution tool, since regularization of deconvolution algorithms already does the same job very efficiently. So this step is optional. Linear masks are however very efficient for noise reduction of linear images. In fact, I have implemented an integrated linear mask generation feature in the MultiscaleLinearTransform and MultiscaleMedianTransform tools that we have recently released. But let's stay on topic.


Step 3: Local Deringing Support

A local deringing support is a special image used by the Deconvolution tool to drive a deringing routine that works at each deconvolution iteration to limit the growth of ringing artifacts. Although a deringing support image looks and is built like a mask, it is important to point out that it is not a mask and does not work as such.

For a deep-sky image, a local deringing support image can be built just as a star mask. In this case, instead of the StarMask tool (which is currently subject to a deep revision), I'll implement a step-by-step procedure based on multiscale analysis. We begin with a stretched duplicate of the image. I have just transferred STF parameters to HistogramTransformation.

(http://forum-images.pixinsight.com/20140424/Deconvolution/08-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/08.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/08.jpg)

The next step is a strong instance of HDRMultiscaleTransform. The purpose of this process is to flatten the image, so jump discontinuities (e.g., stars) can be isolated more easily in subsequent steps.

(http://forum-images.pixinsight.com/20140424/Deconvolution/09-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/09.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/09.jpg)

The starlet transform can be used to remove all large scale structures and high-frequency noise. This is achieved by disabling the residual and first layers, respectively.

(http://forum-images.pixinsight.com/20140424/Deconvolution/10-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/10.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/10.jpg)

A histogram stretch will intensify deringing support structures:

(http://forum-images.pixinsight.com/20140424/Deconvolution/11-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/11.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/11.jpg)

followed by a convolution to make them larger:

(http://forum-images.pixinsight.com/20140424/Deconvolution/12-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/12.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/12.jpg)

Finally, a sequence of histogram stretches and convolutions allows us to achieve the degree of protection required:

(http://forum-images.pixinsight.com/20140424/Deconvolution/13-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/13.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/13.jpg)

Basically, we want to protect bright stars and other high-contrast, small-scale structures, which are the structures where ringing becomes particularly problematic. To control deringing support generation, we can duplicate the image and activate the deringing support as a mask.

(http://forum-images.pixinsight.com/20140424/Deconvolution/14-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/14.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/14.jpg)


Step 4: Deconvolution

We have now everything we need for deconvolution: a linear image with a reasonable amount of signal, a good local deringing support and, as an option, a linear mask that will provide additional protection to low-signal regions.

This is a preview on M82 before deconvolution:

(http://forum-images.pixinsight.com/20140424/Deconvolution/15-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/15.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/15.jpg)

and after deconvolution:

(http://forum-images.pixinsight.com/20140424/Deconvolution/16-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/16.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/16.jpg)

Note that ScreenTransferFunction has to be used to reveal significant structures within the brightest areas of the image, where we want to control how deconvolution is doing its work. This is the same comparison for M81, before deconvolution:

(http://forum-images.pixinsight.com/20140424/Deconvolution/17-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/17.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/17.jpg)

and after deconvolution:

(http://forum-images.pixinsight.com/20140424/Deconvolution/18-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/18.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/18.jpg)

The best way to learn how the most critical deconvolution parameters work is by comparing the results with and without them. This is a more stretched view of the deconvolved M81 preview with regularization:

(http://forum-images.pixinsight.com/20140424/Deconvolution/19-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/19.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/19.jpg)

and without regularization:

(http://forum-images.pixinsight.com/20140424/Deconvolution/20-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/20.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/20.jpg)

Note the noise intensification when we disable deconvolution regularization. In this case I have lowered the regularization threshold for the second wavelet layer; perhaps this has been an error.

This is what happens if we disable deringing:

(http://forum-images.pixinsight.com/20140424/Deconvolution/21-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/21.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/21.jpg)

and this is the result without local deringing:

(http://forum-images.pixinsight.com/20140424/Deconvolution/22-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/22.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/22.jpg)

Global deringing always degrades the result of deconvolution to some degree, so we always must try to find the smallest value of the global dark parameter that prevents ringing. If we use a good local deringing support, we can use an even smaller global dark value because local deringing is very effective to prevent ringing around the brightest stars. The optimal values, as always, depend on the image.

Also take into account that a very small amount of ringing can actually be beneficial because it increases acutance, which leads to a higher visual perception of detail.


Step 5: Nonlinear Stretch

The next processing steps should include some noise reduction on the linear image after deconvolution. For this purpose MultiscaleLinearTransform, MultiscaleMedianTransform (especially the new median-wavelet algorithm) and TGVDenoise are the tools of choice. I'll skip these tasks and will go directly to the nonlinear stretch step. In this case I have just transferred STF to HistogramTransformation.


Step 6: HDR Compression

Some HDR compression will allow us to reveal the result of deconvolution on high signal areas. The HDRMultiscaleTransform tool yields a very nice result for this M81/M82 image:

(http://forum-images.pixinsight.com/20140424/Deconvolution/23-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/23.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/23.jpg)

This is a 1:1 view on M82:

(http://forum-images.pixinsight.com/20140424/Deconvolution/24-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/24.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/24.jpg)

and M81:

(http://forum-images.pixinsight.com/20140424/Deconvolution/25-tn.jpg) (http://forum-images.pixinsight.com/20140424/Deconvolution/25.jpg)
Click for full-size image (http://forum-images.pixinsight.com/20140424/Deconvolution/25.jpg)


I hope this brief tutorial will help you to achieve better results with our deconvolution tools from now on. Let me know if you want more information on the applied processes, or more detailed descriptions of some processing steps.
Title: Re: Deconvolution Challenge
Post by: chris.bailey on 2014 April 25 08:05:39
Juan

Thats a very useful decon walkthrough. One question - I realise the local support image is not a mask but should it be inverted (white background) prior to be used in the deconvolution tool?

Chris
Title: Re: Deconvolution Challenge
Post by: jeffweiss9 on 2014 April 25 14:34:09
That's just great, Juan.  I'll be studying it further in detail and I'm sure it will help a lot of folks.
Thanks very much.
-Jeff
Title: Re: Deconvolution Challenge
Post by: astroedo on 2014 April 27 06:08:25
Juan

Thats a very useful decon walkthrough. One question - I realise the local support image is not a mask but should it be inverted (white background) prior to be used in the deconvolution tool?

Chris

I'm not Juan, but I think that the right answer in NO, you do not have to invert the local deringing support: such a "mask" tells to the deconvolution process how much deringing has to be applied locally.
You need much more deringing around the stars, so stars should be white on a dark background where local deringing is not needed and only global deringing applies.

Is it correct Juan?

bye

Edoardo
Title: Re: Deconvolution Challenge
Post by: Juan Conejero on 2014 April 28 09:20:42
Yes, absolutely correct.
Title: Re: Deconvolution Challenge
Post by: Ignacio on 2014 April 28 10:47:15
Very interesting deconv example, Juan, thank you.

Two things called my attention. First, the "unusual" way of creating a star mask for deringing support, which begs the question: what should we expect from the new/revised  StarMask module?
Second: I typically build a support mask to cover the bigger/brightest stars, but I see in your example that pretty much all stars are supported. On the positive side, I see that this prevents rings around them, but on the negative side they are (apparently) not reduced in size. In my experience with deconv, medium and small size stars are nicely reduced (fwhm cut in half) if not supported, and without darks rings around them. In some cases, when such stars are embedded in a region with even and relatively bright background, then it is harder to come up with the proper deringing parameters if unsupported, and there is always a tradeoff.

Thoughts/comments?

Ignacio
Title: Re: Deconvolution Challenge
Post by: Juan Conejero on 2014 April 28 11:24:13
Thanks Ignacio. The StarMask tool has three well-known problems:

- It is very slow.
- It requires a lot of trial/error work.
- It is not previewable.

These limitations cause StarMask to be difficult to use in many practical cases, even with modern hardware. This obviously has to be addressed with a revision of the tool. If I remember well, I wrote StarMask at least seven years ago, so a complete redesign/reimplementation is in order.

Processing wise, stars and other high-contrast, small-scale structures can be considered as singularities where most image processing algorithms fail or are not applicable. Consequently, having efficient tools to isolate stars is of the highest importance. We already have multiscale analysis tools that, if creatively used, allow for very efficient generation of star masks, and I just wanted to show a practical example in this tutorial. I probably was too exhaustive in the description of these techniques, and hence the generated deringing support is excessive. As you point out, the key word here is to find a compromise between ringing suppression and deconvolution efficiency.
Title: Re: Deconvolution Challenge
Post by: Ignacio on 2014 April 28 11:33:29
Understood, thanks. Looking forward to the new tool.

Ignacio
Title: Re: Deconvolution Challenge
Post by: jeffweiss9 on 2014 April 28 21:33:19
Juan-   Update
  Ok, I worked through it all in detail and found it was a very good tutorial.  I believe I have a good feeling now for what needs to be done to make it work well, although the test will come in applying it on very different data that requires parameters that may be far from these initial values. But I think it gave me a good appreciation of the intermediate goals for each step, independent of the parameters, so I'm hopeful I now will be able to achieve consistent results.
Thanks for the great tutorial.
Clear skies,
-Jeff
Title: Re: Deconvolution Challenge
Post by: Carlos Milovic on 2014 July 07 15:19:40
Just a quick note on the news from the development team:
- Our first implementation of TGVRestoration was modeled with a Gaussian noise distribution as basis. So, it worked very well with  high SNR images, and "normal" images. Astronomical deep sky images, on the other hand, suffered from some serious ringing. We included some deringing functions to mitigate this problem, and achieve more robust results, but were not as good as I expected (and I had very high expectations for this... I think that it is better than regularized RL, but more work is needed).
- The tests on deep sky images derived in a reformulation of the TGV regularization algorithm, both for the Denoise and Restoration problems. A new method was designed for a Poisson noise model. Right now we are close to publish a new TGVDenoise, with several changes. Now the tool is much more flexible, and may adapt its behavior for many noise distributions. It has in-built 3 statistical noise models, Gaussian, Poisson and a L1 norm, with a new flexible edge protection. We are polishing the interface elements and working on some examples to accompany the new release.
- The development of the TGVRestoration tool has being delayed a couple of weeks, until we have the new TGVDenoise ready for release. The new TGVR will also incorporate several statistical models. Right now we are testing a mixed L1/L2 Norm (for Gaussian noise), and three solvers for the Poisson model. One of them is a regularized Richardson-Lucy iteration, with a modified gradient, following the classic TV algorithm. Another is a two step Expectation-Maximization (similar to regularized RL), with a TGV step. The last is a new derivation of the Chambolle/Pock primal dual algorithm for Poisson noise. So, a lot is going on here.

Please, have a little more patience. We are working to create a very powerful and flexible tool. We still need to evaluate the efficiency of our new deringing methods, so more time is needed. I may publish an unofficial release in my development server, asking for beta testing. Stay tuned.

Title: Re: Deconvolution Challenge
Post by: Philippe B. on 2014 October 27 09:47:08
Hello PI team  O0


Should incoming 1.8.3 see the new TGVRestoration process ?


Best
Philippe
Title: Re: Deconvolution Challenge
Post by: Carlos Milovic on 2014 October 28 01:06:52
Hi Phillipe.

Right now we are working on the last modifications to TGVDenoise. There are a lot of changes. We are evaluating some critical changes to the interfase and inner working of some parameters.
Also we have the inpainting tool based on tgv that needs a small review.
Sideways, I wrote the first documentation of TGVDenoise, for the current release. It may be online in the next days or couple of weeks.

Development of TGVRestoration should continue this december. I made huge changes to the code also, that needs a lot of testing. Unfortunatelly, I'm quite busy with other projects right now, so it will have to wait a little while.


Thanks for your patience! I will consider to release an alpha version of TGVR if I see that it works without problems in the first tests.
Title: Re: Deconvolution Challenge
Post by: Philippe B. on 2014 October 29 04:28:55
Hi Carlos
Thank you for your work ! I'm sure new TGVDenoise will be very nice and I will wait for TGVR (but if you want I test some alpha version, please do not hesitate)


Cheers