DirectionalField

DirectionalField is a filter that generates a directional field representation of an input image, utilizing operations such as thresholding, contour detection, jump flooding, and direction decoding. The generated output encodes directional and distance information from the contours of the input.

The filter supports a variety of configurable properties such as thresholds, scaling, and different modes for direction and magnitude representation.

This class extends Filter1to1, processing one input ColorBuffer and producing one output ColorBuffer.

Parameters:

• threshold: The threshold value used during the binary segmentation of the input image.

• distanceScale: The scale factor applied to the distance values encoded in the output.

• normalizedDistance: Whether to normalize the distance values in the output.

• unitDirection: Whether to represent gradient directions as unit vectors.

• signedMagnitude: Whether to encode magnitude with signed values.

• flipV: Whether to flip the vertical component of the direction vectors in the output.

Lifecycle:

• Resources such as intermediate ColorBuffer instances are created dynamically based on the dimensions of the input image. These resources are cleaned up in the destroy method to prevent memory leaks.

Responsibilities:

• Threshold the input to create a binary image.

• Detect contours from the thresholded image.

• Generate a jump flood field to calculate distance and direction information.

• Decode directional data into the final output.