Problem in compiling mex file

[jM]

Hi.

I want to use and install HDR VDP metric.

I have followed the procedure outlined in read me file and have downloaded matlabpyrtools fixed version and added it into matlab path along with hdrvdp toolbox . As it is mentioned in read me file that that it is important to compile the matlabpyrtool for proper working of metric, when i try to run compilePyrTools for compiling after adding path in matlab, it gives me an error. 

C:\PROGRA~1\MATLAB\R2012A\BIN\MEX.PL: Error: 'upConv.c' not found. 

 

Error using mex (line 206)

Unable to complete successfully.

Error in compilePyrTools (line 10)

mex upConv.c convolve.c wrap.c edges.c

I have got the same error when i tried to install on linux. 

It does show me results when i try following example without compiling  the mex file 

reference = logspace( log10(0.1), log10(1000), 512 )' * ones( 1, 512 );

test = reference .* (1 + (rand( 512, 512 )-0.5)*0.1);

res = hdrvdp( test, reference, 'luminance', 30 );

clf;

Any suggestion, comment on issue will be highly appreciated.

Thank you for your time

[Rafal]

On Tuesday, 13 May 2014 14:35:16 UTC+1, jM wrote:

Hi.

I want to use and install HDR VDP metric.

I have followed the procedure outlined in read me file and have downloaded matlabpyrtools fixed version and added it into matlab path along with hdrvdp toolbox . As it is mentioned in read me file that that it is important to compile the matlabpyrtool for proper working of metric, when i try to run compilePyrTools for compiling after adding path in matlab, it gives me an error. 

C:\PROGRA~1\MATLAB\R2012A\BIN\MEX.PL: Error: 'upConv.c' not found. 

Make sure that you execute compilePyrTools.m while your current directory is "MEX". 

 

I have got the same error when i tried to install on linux. 

It does show me results when i try following example without compiling  the mex file 

reference = logspace( log10(0.1), log10(1000), 512 )' * ones( 1, 512 );

test = reference .* (1 + (rand( 512, 512 )-0.5)*0.1);

res = hdrvdp( test, reference, 'luminance', 30 );

clf;

For some reason MEX and  matlab version of the steerable pyramid decomposition returns different results. HDR-VDP-2 was calibrated and tested using the MEX functions. PyrTools will work without the MEX files, but HDR-VDP-2 will be returning wrong results.

Rafal

 

[jM]

Dear Rafal.

Many thanks for your response. It seems I misunderstood README as I was adding path of MEX folder in matlab instead of changing my current working directory to MEX folder.

In HDR VDP while using vdpvis there was an option of writing summary. The ratio of pixels that contained distortion of probability greater than 75% and 95% was written in that summary. Is this possible in HDR-VDP2 or one has to calculate it using P_maps and does P_det gives maximum probability of detection ?

Thank you

[Rafal]

On Wednesday, 14 May 2014 18:53:40 UTC+1, jM wrote:

In HDR VDP while using vdpvis there was an option of writing summary. The ratio of pixels that contained distortion of probability greater than 75% and 95% was written in that summary. Is this possible in HDR-VDP2 or one has to calculate it using P_maps and does P_det gives maximum probability of detection ?

 

No, there is no such an option. These values can be easily computed from P_map. However, it is recommended for HDR-VDP-2 to use "Q" value instead (for quality predictions). An alternative is to use C_map and a pooling function of your choice (Minkowski summation).

Rafal

[jM]

Thank you Rafal.

Regards