integralKernel
Define filter for use with integral images
Description
An integralKernel
object describes box filters for use with
integral images.
Creation
Description
intKernel = integralKernel(bbox,weights)
creates an upright box
filter from bounding boxes, bbox
, and their corresponding weights,
weights
. The bounding boxes set the BoundingBoxes
property and the weights set the Weights property.
For example, a conventional filter with the coefficients:
and two regions:
region 1: x=1, y=1, width = 4, height = 2 |
region 2: x=1, y=3, width = 4, height = 2 |
boxH = integralKernel([1 1 4 2; 1 3 4 2],[1, -1])
intKernel = integralKernel(bbox,weights,orientation)
creates a
box filter with an upright or rotated orientation. The specified orientation sets the
Orientation property.
Properties
Usage
Computing an Integral Image and Using it for Filtering with Box Filters
The integralImage
function together with the
integralKernel
object and integralFilter
function complete the workflow for box filtering based on
integral images. You can use this workflow for filtering with box filters.
Use the
integralImage
function to compute the integral imagesUse the
integralFilter
function for filteringUse the
integralKernel
object to define box filters
The integralKernel
object allows you to
transpose the filter. You can use this to aim a directional filter. For example, you can
turn a horizontal edge detector into vertical edge detector.
Object Functions
Examples
References
[1] Viola, Paul, and Michael J. Jones. “Rapid Object Detection using a Boosted Cascade of Simple Features”. Proceedings of the 2001 IEEE® Computer Society Conference on Computer Vision and Pattern Recognition. Vol. 1, 2001, pp. 511–518.
Version History
Introduced in R2012a
See Also
detectMSERFeatures
| integralImage
| integralFilter
| detectSURFFeatures
| SURFPoints