Extracts features from image

features = cv.StructuredEdgeDetection.getFeatures(src, opts)

Input

• src source image to extract features (RGB float in [0;1]).
• opts a scalar struct of random forest options (feature params), with the following fields:
  • normRad gradientNormalizationRadius gradient normalization radius.
  • grdSmooth gradientSmoothingRadius radius for smoothing of gradients (using convolution with triangle filter).
  • shrink shrinkNumber amount to shrink channels.
  • nChns numberOfOutputChannels number of edge orientation bins for output.
  • nOrients numberOfGradientOrientations number of orientations per gradient scale.

Output

• features extracted features.

Extracted features are appropriate for StructuredEdgeDetection training.