How do I change input values in different function file ?

I have a .m file with several function file and my focus is to observe the changes in PSNR and SSIM results values. How do I make changes to the INPUT values of any of the function files ? Please guide me. Thank you

[Walter Roberson]

Do you mean that you have calls to input() in the functions to get data from the user, but you want to somehow enter values on the user's behalf as if they had typed in those values?

[Chidiebere Ike]

No. I don't need any data input from the user. The inputs parameters has been set in the function files but can't seem to locate it. For example one of the function file as stated below says:

function [imgs, midres] = scaleup_ANR(conf, imgs)

Parameters has been set for conf input.

How do I locate these parameters and make changes.

That's my question.

Thanks

[Walter Roberson]

I find three git repositories for that code. Which one are you using? The most complete version appears to be at https://github.com/yulunzhang/CRC-SISR.git

[Chidiebere Ike]

Not exactly but the code I am working on is almost similar to the above you shared. I intend to make some modifications in the input parameters but can't seem to locate the parameter from the code. Please guide me to achieve this using the above as reference or example. Thanks so much...

[Chidiebere Ike]

Hello Walter, I am using the code on https://github.com/junjun-jiang/LANR-NLM. I intend to make some modifications to improve the algorithm. Can you please suggest to me an idea to improve the result of this code as this will to help improve my knowledge? Thanks so much

Answer 1

[Walter Roberson]

scaleup_ICR does not appear to be called by the demo code, but scaleup_ICR_MIKSVD is called and expects the same kind of first input.

If you look at line 207 you will see that DEMO_CRC loads conf_ICR_App_1 from a .mat file if the .mat exists, and otherwise builds a struct field by field, starting around line 213. Around line 691 the structure is adjusted to current information and then scaleup_ICR_MIKSVD is called on it.

So you can be fairly sure that scaleup_ICR should be called with a struct that has much the same fields as are constructed for conf_ICR_App_1

[Chidiebere Ike]

Hello Walter, I am using the code on https://github.com/junjun-jiang/LANR-NLM. I intend to make some modifications to improve the algorithm. Can you please suggest to me an idea to improve the result of this code as this will to help improve my knowledge? Thanks so much

[Walter Roberson]

I am not familiar with the theory being used, and I am not familiar with the software package. I am not sure I had ever encountered it before you asked about it and I searched the internet to try to figure out what you were talking about.

You will need to ask a much more focused question. For example if you had profiled the code and found a particular section that was slow, then potentially you could link to your data files and the section of code and ask for suggestions on improving the performance of that section. But I am not going to go read the papers and try to come up with an improvement to the theory.

[Chidiebere Ike]

Thanks Walter for your time and efforts, I do appreciate. My focus is to achieve better Super resolution result with a higher PSNR and SSIM results values. To achieve this, first I came up with an idea to set a Function file to improve the filtering (using NonLocal Means) process of the algorithm and preserve image structure. For example, On Demo_LANR_FAST_NLM_II, line 164, I tried to set a function file named FAST_NLM_II. Also, On "Proposed_NLmeansfilter code", Line 146-153 and Line 211-218; I tried to set a function file but didn't get a better result. Attached is the full codes and its function files. Base on the above, I seek assistance on improving the performance of the filtering process. I will appreciate your guide. Thank you

[Walter Roberson]

Sorry, that is a question of algorithm rather than of MATLAB; I do not think I have anything to contribute to that.

[Chidiebere Ike]

It's Okay, Thanks so much, I appreciate