image =	harrisCorner3dImg (inImg3d)
image =	harrisCorner3dImg (inImg3d,inGradStdDev)

Detailed Description

Computes the Harris corner detection response on a 3d image.

This algorithm is the 3d extension of the Harris corner detection 2d image algorithm [1] proposed by I. Laptev [2] and computes the cornerness for each pixel of the input 3d image. This algorithm also support a extension of Shi-Tomasi [3] [4] corner detector algorithm to 3d case.

As an extension of 2d case, $M$ is defined as follows :

$M = \sum_{x, y, z \in \aleph}{\omega(x, y, z) \begin{bmatrix} I_x^2 & IxIy & IxIz\\ IxIy & Iy^2 & IyIz\\ IxIz & IyIz & Iz^2 \end{bmatrix}}$

With $Ix$ , $Iy$ and $Iz$ are the spatial derivatives of the input image along the directions $x$ , $y$ , and $z$ respectively and $\omega(x, y, z)$ is a Gaussian weight in the neighbourhood $\aleph$ .

If selected method is set to ipsdk::imaproc::attr::eCornerDetectionMethod::eCDM_Harris (which is the default value) the cornerness is calculated at the position by :
$C(u, v, w) = det(M) - k \left( trace(M) \right)^3$
Where is the sensitivity corresponding to the parameter , is the determinant of and is the trace of .
In order to fit into the output image range, the algorithm returns the cube root of :
$OutImg(u, v, w) = \sqrt[\leftroot{3}3]{C(u, v, w)}$
If selected method is set to ipsdk::imaproc::attr::eCornerDetectionMethod::eCDM_ShiTomasi the cornerness can be computed at the position by :
$C(u, v, w) = OutImg(u, v, w) = \min(\lambda_1, \lambda_2, \lambda_3)$
where $\lambda_1$ , $\lambda_2$ and $\lambda_3$ correspond to eigen values of matrix (in this case sensitivity contained by parameter is ignored).

See Harris corner detection 2d image for more details on the algorithm and for an illustration of the Harris corner detection algorithm in the 2d case.

[1] C. Harris and M. Stephens (1988). "A combined corner and edge detector". Proceedings of the 4th Alvey Vision Conference. pp. 147 - 151.
[2] I. Laptev. "On space-time interest points". IJCV , 64(2-3):107 - 123, 2005.
[3] J. Shi and C. Tomasi (June 1994). "Good Features to Track". 9th IEEE Conference on Computer Vision and Pattern Recognition. Springer
[4] C. Tomasi and T. Kanade (2004). "Detection and Tracking of Point Features". Pattern Recognition. 37: 165 - 168

Example of Python code :

Example imports

import PyIPSDK

import PyIPSDK.IPSDKIPLFeatureDetection as fd

Code Example

# Harris corner detection

outImg = fd.harrisCorner3dImg(inImg)

Example of C++ code :

Example informations

Header file

#include <IPSDKIPL/IPSDKIPLFeatureDetection/Processor/HarrisCorner3dImg/HarrisCorner3dImg.h>

Code Example

    // Sample with a generated output image
    // ------------------------------------
    ImagePtr pAutoOutImg = harrisCorner3dImg(pInImg);
    // Sample with a provided output image
    // ------------------------------------
    // create output image
    const ipUInt64 sizeX = pInImg->getSizeX();
    const ipUInt64 sizeY = pInImg->getSizeY();
    const ipUInt64 sizeZ = pInImg->getSizeZ();
    ImageGeometryPtr pOutputImageGeometry = geometry3d(outType, sizeX, sizeY, sizeZ);
    boost::shared_ptr<MemoryImage> pOutImg(boost::make_shared<MemoryImage>());
    pOutImg->init(*pOutputImageGeometry);
    // compute harris image
    harrisCorner3dImg(pInImg, stdDev, createGaussianCoverage(gaussRatio, 2), createHarrisParams(sensitivity), pOutImg);