Segmenting and masking all shades of red from an image using opencv

I am trying to segment all shades of red form an image using hue saturation values and use InRangeS function to create a mask which should have all red areas whitened and all others blacked(a new 1 channel image). Thwn Inpaint them to kind of obscure the segmented portions. My code is as given. However I am unable to get an output image, it doesnt segment the desired color range.

Any pointers on my approach and why it isnt working. ?

int main(){
IplImage *img1=cvLoadImage("/home/techrascal/projects/test1/image2.jpeg");
//IplImage *img3;
IplImage *imghsv;
IplImage *img4;
CvSize sz=cvGetSize(img1);

imghsv=cvCreateImage(sz,IPL_DEPTH_8U,3);
img4=cvCreateImage(sz,IPL_DEPTH_8U,1);
int width = img1->width;
int height = img1->height;
int bpp = img1->nChannels;
//int w=img4->width;
//int h=img4->height;
//int bn=img4->nChannels;
cvNamedWindow("original", 1);
cvNamedWindow("hsv",1);
cvNamedWindow("Blurred",1);
int r,g,b;




// create inpaint mask: img 4 will behave as mask
cvCvtColor(img1,imghsv,CV_BGR2HSV);
CvScalar  hsv_min = cvScalar(0, 0, 0, 0);
CvScalar  hsv_max = cvScalar(255, 0, 0, 0);
//cvShowImage("hsv",imghsv);
cvInRangeS( imghsv, hsv_min, hsv_max, img4 );
cvInpaint(img1, img4, img1, 3,CV_INPAINT_NS );
cvShowImage("Blurred",img1);
    cvReleaseImage(&img1);
    cvReleaseImage(&imghsv);
cvReleaseImage(&img4);
//cvReleaseImage(&img3);
char d开发者_开发百科=cvWaitKey(10000);
cvDestroyAllWindows();

    return 0;}

Your code logic seems correct but you will definetely need to adjust your hsv range values (hsv_min and hsv_max).

Read this detailed guide that show you hsv range defined in opencv http://www.shervinemami.co.cc/colorConversion.html