COMPILATIONS ERROR using DEV C++ to compilate a image process program

luisfne · 11-09-2011, 04:03 PM

I have a problem when I try to compilate this prog...I'm trying to process images to generate a color distortion image model ...It's just the beggining of my main program, but if I don't solve this...

my program wont work...plz help me!!

ERRORs:

1)Error:
"167 C:\OCV\DistorçãoCoreBrilho.cpp conversion from `double' to non-scalar type `CvScalar' requested "

""

double normaquadframe,dotquad;
double quada,normaquadref;
double g,colordist;
int i, j;

while(1)
{
CvScalar frameM, refM;//,colordist;//colordistM,colordist,
for(int j=0;j<frame->height;j++)
for(int i=0;i<frame-> width;i++)

{

frameM = cvGet2D(frame,i,j);
refM = cvGet2D(ref,i,j);

normaquadframe=(pow(frameM.val[0],2)+pow(frameM.val[1],2)+pow(frameM.val[2],2));

normaquadref=(pow(refM.val[0],2)+pow(refM.val[1],2)+pow(refM.val[1],2));

dotquad=pow(((frameM.val[0]*refM.val[0])+(frameM.val[1]*refM.val[1])+(frameM.val[2]*refM.val[2])),2);

// normaquadframe*pow(cosf(c/(a*b),2)

quada= (dotquad/normaquadref);
g=(normaquadframe-quada);

colordist=(sqrt(g));
167 cvSet2D(colordistM,i,j,colordist);
}

2)
Error:
"166 C:\OCV\DistorçãoCoreBrilho.cpp no match for 'operator=' in 'colordist = sqrt(g)' "

" note C:\OpenCV2.2\include\opencv2\core\types_c.h:1199 candidates are: CvScalar& CvScalar:

perator=(const CvScalar&) "

double normaquadframe,dotquad;//,quada;//,colordist;
double quada,normaquadref;//=0.0001, colordist=0.0001;
double g;//,colordist;
int i, j;

while(1)
{
CvScalar frameM, refM,colordist;//colordistM,colordist,
for(int j=0;j<frame->height;j++)
for(int i=0;i<frame-> width;i++)

{

frameM = cvGet2D(frame,i,j);
refM = cvGet2D(ref,i,j);

normaquadframe=(pow(frameM.val[0],2)+pow(frameM.val[1],2)+pow(frameM.val[2],2));

normaquadref=(pow(refM.val[0],2)+pow(refM.val[1],2)+pow(refM.val[1],2));

dotquad=pow(((frameM.val[0]*refM.val[0])+(frameM.val[1]*refM.val[1])+(frameM.val[2]*refM.val[2])),2);

// normaquadframe*pow(cosf(c/(a*b),2)

quada= (dotquad/normaquadref);
g=(normaquadframe-quada);

166 colordist=(sqrt(g));
cvSet2D(colordistM,i,j,colordist);
}

jhwilliams · 11-09-2011, 09:25 PM

Hi Luis,

It's impossible to read this! Could you please edit your post to use [ code ] [ /code ] tags (no spaces)? It'd be easier to help.

Best,
Jameson

firstfire · 11-09-2011, 10:19 PM

Hi.
According to this link your `colordist' (pixel color) variable should be of type CvScalar, not double. To set the actual color, you should use something like

Code:

colordist.val[0]=111;
colordist.val[1]=111; // for more than 1-channel image
colordist.val[2]=111; // ditto

You should decide how resulting pixel color (in general up to four channels per pixel) related with your scalar values like `sqrt(g)' etc and construct the pixel correspondingly.

luisfne · 11-10-2011, 06:10 AM

Quote:

Originally Posted by firstfire

Hi.
According to this link your `colordist' (pixel color) variable should be of type CvScalar, not double. To set the actual color, you should use something like

Code:

colordist.val[0]=111;
colordist.val[1]=111; // for more than 1-channel image
colordist.val[2]=111; // ditto

You should decide how resulting pixel color (in general up to four channels per pixel) related with your scalar values like `sqrt(g)' etc and construct the pixel correspondingly.

This information will be helpfull to me!!! I'll try to do this code by another way ...and I'll confirm to you if worked better ..Thank you so much firstfire!!