Download

krawtchouk.zip - August 2023

Once the download is completed, copy the downloaded .zip file to your working directory and unzip it.

Input Arguments and Output

The core files of our package include readImage.m and get3DKMI.m. The function get3DKMI plays a crucial role in receiving user input and producing the final output. Therefore, understanding the formats of inputs and outputs is key to helping users achieve the desired outcomes effectively.

Input	Description
arr (Mandatory)	An array representing the 3D image. The image should be binarized, i.e., the pixel intensity at any point on the object is 1, otherwise, it is 0.
dim (Mandatory)	An intended size of the image. To maintain the spherical coverage shape of the weight function when extracting the local features, we require the 3D image to be a cube. It is important to use a consistent size when comparing the features extracted from different 3D images.
order (Optional)	An integer that specifies the moment order. A higher value indicates a more comprehensive description. Nonetheless, a larger number necessitates increased memory and computational time. Default value: 2
POI (Optional)	"POI" stands for "point-of-interest", denoting a 1x3 array that specifies the coordinates of the point of interest within the image. If the user specifies a POI as input, then the localized features around the POI will be extracted, whereas the absence of this input results in the extraction of features from the image as a whole. Default value: []

Output	Description
invariants	An 1xP array representing the extracted features (3D Krawtchouk moment invariants). The length P of the features depends on the order.