Integration "Large scale pixel rejection" rejects nothing?

[UlteriorModem]

Greetings,

I have been playing with different debayer methods and channel combination to get best results from a Duo Band narrow band filter and OSC. Which is kind of irrelevant but...

I go to integrate the registered results from WBPP and select both high and low large scale pixel rejection but it does not reject anything. All are 0.00 ?

No rejection by Tom Whit, on Flickr

[UlteriorModem]

Hrm, 57 views and nothing?

btw I tried playing with the sliders for high and low rejection. Both high and low settings. from 1 to 5 I think and still nothing rejected.

When I look at rejection maps when I let WBPP do the integration there is plenty, maybe even too many rejections.

[Harry page]

Hi

Perhapes you do not have any large scale items to remove ( sat trails etc )

I presume you have selected a pixel rejection method and sigma under pixel rejection 1 and 2 , I suspect this is the problem

Harry

[ngc1535]

Just to expand on what has been said so far, Large Scale Rejection operates on the map of rejected pixels (not on the images themselves). The "structures" that are referred to are the rejected things that the rejection methods find based on the statistical analysis.

So Large Scale Rejection will not affect an empty map- the 0 rejection you see is due to the fact your rejection settings (under (1) and (2) tabs) are not rejecting anything.
With only 6 observations you will be quite limited in the amount you can reject based on any algorithm. Percentile rejection will certainly be a good place to go.

Once you have rejected things using a rejection algorithm... then turning on Large Scale rejection will do something. It will make larger (based on the scales sizes you choose) the detected/rejected bits of the image. This can help make certain anything that is determined to be an outlier- may have associated neighboring pixels which are also really outliers- but due to their low S/N can't be eliminated without the help of this Large Scale rejection method. Most times your rejected pixels will completely characterize the things you need to be rejected...and no modification of the map (Large Scale rejection) is required.

It really is an awesome tool when it is necessary to use.

-adam

[UlteriorModem]

Thank you for the input. Exactly what I was looking for.

I did not really understand the manner in which it worked. I will try the suggestions and get back to you.

Well that did not take long. I changed to percentile clipping as recommended and poof now it works

Do these results look right? Something must be right because it did a grand job of nullifying the satellite trail obvious in the high rejection map.

Rejection 1 by Tom Whit, on Flickr