API

The API is accessed through the #include <chomravec/chromavec.h> header. All of the utilities and functions reside within the chromavec:: namespace.

Filtering Functions

enum GradientMode

Enums that define the set of output modes for the ColourVectorGradientFilter().

enum kDirectOutput

Output a raw gradient image that is identical to what is passed internally within the chromavec library.

enum kMagnitudeOnly

Output the gradient magnitude image, scaled to be within 0 to 255.

enum kToHSV

Output the gradient as an RGB image with HSV colouring. The gradient magnitude will be the value while the angle will be in the hue.

cv::Mat VectorMedianFilter(const cv::Mat &img, const int window)

Applies the Vector Median Filter onto an image. This is a noise reduction filter that is quite similar to the standard median filter.

Parameters:
  • img – input image
  • window – filtering window size
Returns:

filtered image
cv::Mat VectorRangeFilter(const cv::Mat &img, const int window = 5)

Applies the Vector Range Filter onto an image. The vector range filter is a type of edge detector that returns the distance between the most and least central RGB vector in a window.

Parameters:
  • img – input image
  • window – filtering window size
Returns:

filter response map
cv::Mat MinimumVectorDispersionFilter(const cv::Mat &img, const int k = 3, const int l = 4, const int window = 5)

The Minimum Vector Dispersion Filter is a combination between the VMF and VRF filters. The output of the filter is the distance between the average k most-central vectors and the l least-central vectors. This has a noise reduction affect on the edge map so that the edge response is less sensitive to noise in the image.

Parameters:
  • img – input image
  • k, l – the two parameters used to control between noise suppression and edge detection
  • window – filtering window size
Returns:

filter response map
cv::Mat ColourVectorGradientFilter(const cv::Mat &img, const double sigma = 0, const GradientMode mode = kToHSV)
Parameters:
  • img – input image
  • sigma – the sigma of a Gaussian pre-filter
  • mode – the gradient output mode
Returns:

colour gradient image
cv::Mat ColourCannyEdgeDetect(const cv::Mat &img, const double t1, const double t2, const double sigma = 3.0)

Perform Canny-style edge detection using colour gradients.

Parameters:
  • img – input image
  • t1, t2 – the lower and upper Canny hysteresis thresholds
  • sigma – pre-blurring amount

Miscellaneous

The library version information is stored within a data structure, documented below, that is generated at build time. This is automatically included in the main library header.

class Version

The Version struct provides version information about the chromavec library. The values that it provides are filled in at build time and they follow semver convention.

static constexpr int major

Major version number.

static constexpr int minor

Minor version number.

static constexpr int patch

Patch version number.

static std::string ToString()

Returns the version information as a string. The string will be in <major>.<minor>.<patch> form.