Histogram of Oriented Gradients (HOG)

Brief Description

File: \examples\Processing\Geometry\ObjectFeatures\HOG\HOG_4Bins_HistogramMax.vad
\examples\Processing\Geometry\ObjectFeatures\HOG\HOG_9Bins_HistogramMax.vad
\examples\Processing\Geometry\ObjectFeatures\HOG\HOG_9Bins_Histogram.vad

Default Platform: imaFlex CXP-12 Quad

Short Description

Calculation of Histogram of Oriented Gradients (HOG)

The VisualApplets design examples "HOG_4Bins_HistogramMax.vad", "HOG_9Bins_HistogramMax.vad" and "HOG_9Bins_Histogram.vad"calculate the Histogram of Oriented Gradients (HOG) for a grayscale image with maximum dimensions of 1024x1024 pixels.The algorithm is mainly based on the publication of [Dal05]. The HOG is a feature descriptor and can be used for object recognition. In the following subsections first the theory is described before the applet design is introduced.