DRAW_SEGMENTATION

DrawSegmentationStep is a pipeline step that configures how to draw a segmentation mask, optionally on an image.
Configs
Descriptions
classColors
This is an optional field which specifies the list of colors to use for each class. The color can be a hex or HTML string like #788E87, an RGB value like - rgb(128,0,255) or  it can be from a set of predefined HTML color names: white, silver, gray, black, red, maroon, yellow, olive, lime, green, aqua, teal, blue, navy, fuchsia, purple.
segmentArray
Name of the NDArray with the class indices, 0 to numClasses-1. In shape of [1, height, width].
image
An optional field, specifying the name of the image to draw the segmentation classes onto. If not provided, the segmentation classes are drawn onto a black background image.
outputName
Name of the output image. Default is named as image.
opacity
An optional field, that is used only used when image key name is set. This specifies, the opacity, between 0.0 and 1.0, of the mask to draw on the image. Default value of 0.5 is used if it's not set. Value of 0.0 is fully transparent and 1.0 is fully opaque.
backgroundClass
An optional field, specifying a class that's not to be drawn. If not set, all classes will be drawn.
imageToNDArrayConfig
Used to account for the fact that n-dimensional array from ImageToNDArrayConfig may be used to crop images before passing to the network, when the image aspect ratio does not match the NDArray aspect ratio. This allows the step to determine the subset of the image actually passed to the network that produced the segmentation prediction to be drawn. Refer config from IMAGE_TO_NDARRAY.

DRAW_POINTS

EXTRACT_BOUNDING_BOX

Last updated 3 months ago

Configs	Descriptions
classColors	This is an optional field which specifies the list of colors to use for each class. The color can be a hex or HTML string like `#788E87`, an `RGB` value like - `rgb(128,0,255)` or it can be from a set of predefined HTML color names: `white`, `silver`, `gray`, `black`, `red`, `maroon`, `yellow`, `olive`, `lime`, `green`, `aqua`, `teal`, `blue`, `navy`, `fuchsia`, `purple`.
segmentArray	Name of the NDArray with the class indices, 0 to numClasses-1. In shape of `[1, height, width]`.
image	An optional field, specifying the name of the image to draw the segmentation classes onto. If not provided, the segmentation classes are drawn onto a black background image.
outputName	Name of the output image. Default is named as `image`.
opacity	An optional field, that is used only used when `image` key name is set. This specifies, the opacity, between `0.0` and `1.0`, of the mask to draw on the image. Default value of `0.5` is used if it's not set. Value of 0.0 is fully transparent and 1.0 is fully opaque.
backgroundClass	An optional field, specifying a class that's not to be drawn. If not set, all classes will be drawn.
imageToNDArrayConfig	Used to account for the fact that n-dimensional array from `ImageToNDArrayConfig` may be used to crop images before passing to the network, when the image aspect ratio does not match the NDArray aspect ratio. This allows the step to determine the subset of the image actually passed to the network that produced the segmentation prediction to be drawn. Refer `config` from IMAGE_TO_NDARRAY.