Convert any image to colorable while creating closed areas with OpenCV and iOS

Question

I'm playing around with iOS and OpenCV trying to create a game which converts basically any image to a black and white colorable image where there are multiple closed areas and we can color these areas with only a tap on the screen. So far I have partial success. The conversion works fine, but I'm stuck at creating an image with concrete colorable (closed) areas.

Is it even possible? If not what alternative methods/ideas do you suggest? I have two, but I like neither.
  - we define areas by drawing after the conversion
  - the tap on the screen only does the recursive search for x pixels

Example: Dropbox pics

Any constructive help is greatly appreciated!

My view controller's related part:

- (void)viewDidLoad{
    [super viewDidLoad];

    imageProcessor = [[Conversion alloc] init];

    _image = [UIImage imageNamed:@"lenna"];
    imageProcessor.thresholdSlider = self.thresholdSlider;
    imageProcessor.thresholdSlider2 = self.thresholdSlider2;

    self.imageView.image = [imageProcessor convertImage:_image]; 
}

- (IBAction)sliderChanged:(id)sender {
    self.imageView.image = [imageProcessor convertImage:_image];
}

And the Conversion class's related part using OpenCV

- (UIImage*)convertImage:(UIImage*)image{

    cv::Mat src;
    cv::Mat src_gray;

    //Convert UIImage to cv::Mat
    src = [self cvMatFromUIImage:image];

    if (!src.empty()) {
        //Convert cv::Mat to grayscale
        cv::cvtColor(src, src_gray, CV_RGBA2GRAY);
        cv::blur(src_gray, src_gray, cv::Size(3,3));

        cv::Mat canny_output;
        cv::vector<cv::vector<cv::Point> > contours;
        cv::vector<cv::Vec4i> hierarchy;

        int thresh = (int)self.thresholdSlider.value;
        int thresh2 = (int)self.thresholdSlider2.value;

        NSLog(@"%d, %d", thresh, thresh2);

        //Detect edges using canny
        //http://en.wikipedia.org/wiki/Canny_edge_detector
        cv::Canny(src_gray, canny_output, thresh, thresh2, 3, true);

        //InputOutputArray image, OutputArrayOfArrays contours, OutputArray hierarchy, int mode, int method,
        //Point offset = Point())
        cv::findContours(canny_output, contours, hierarchy, cv::RETR_TREE, cv::CHAIN_APPROX_SIMPLE, cv::Point(0,0));

        cv::Scalar blackColor = cv::Scalar::all(0);
        cv::Scalar whiteColor = cv::Scalar::all(255);

        //create a WHITE Mat with the canny_output size
        cv::Mat drawing(canny_output.size(), CV_8UC3, whiteColor);
        for (size_t i = 0; i < contours.size(); i++) {
            //InputOutputArray image, InputArrayOfArrays contours, int contourIdx, const Scalar &color, int thickness = 1,
            //int lineType = 8, InputArray hierarchy = noArray(), int maxLevel = 2147483647, Point offset = Point())
            cv::drawContours(drawing, contours, int(i), blackColor, 1, CV_AA, hierarchy, INT_MAX, cv::Point());
        }
        return [self UIImageFromCVMat:drawing];

    }
    return nil;
}

Answer 1

You can follow this steps,

1. Find contour.

2. For each contour check whether its closed or not as explained here, and then draw the contour with appropriate color.