🟠 Extract Images from Mask List

Function Description

This operator extracts (or "crops out") regions corresponding to each mask from a source image based on one or more input binary masks. You can control the extraction shape, such as using precise mask contours, using their bounding rectangles, or extracting only target edges.

Usage Scenarios

After obtaining target masks through segmentation or detection algorithms, this operator can individually extract the image regions of each target, for example removing complex background information and displaying only regions of interest.

Input Output

Input

Image: A source image to be extracted from.

Detection result mask image list: A list composed of one or more binary mask images. Each mask defines a region to be extracted from the source image.

Output

Masked image list: An image list where each image is the result of the source image extracted by the corresponding mask.

Parameter Description

Minimum Bounding Box

Parameter Description

Extract point clouds based on the minimum tilted bounding box of detection instances. Suitable for scenes with relatively regular object shapes, capable of boxing targets with the smallest rectangle.

Parameter Adjustment

  • Bounding box expansion pixels: Specifies pixel values for outward or inward expansion based on the minimum bounding box. Positive values will expand the extraction range, helping to include more information outside object edges; negative values will shrink the range, helping to remove edge noise.

  • Enable reverse: Reverses extraction logic. Disable means extracting point clouds inside the specified shape; enable can be used to remove identified objects and retain background.

Mask

Parameter Description

Extract point clouds based on pixel-level masks of detection instances. Higher precision, capable of precisely extracting point clouds of arbitrary shape objects, but with relatively larger computational load.

Parameter Adjustment

  • Extract mask edges only: Used to analyze object contour features. When enabled, the operator only extracts point clouds at mask contour edges rather than the entire mask interior.

  • Edge thickness: Sets the pixel width of edge contours. Larger values result in thicker extracted edge point cloud bands containing more points.

  • Enable reverse: Reverses extraction logic. Disable means extracting point clouds inside the specified mask; enable can be used to remove identified objects and retain background.

Min Max Values

Parameter Description

Extract point clouds based on axis-aligned bounding boxes (horizontal and vertical rectangles) of detection instances. Fastest computation speed, suitable for scenarios with low precision requirements for extraction regions.

Parameter Adjustment

  • Bounding box X/Y direction expansion pixels: Specifies pixel values for expanding axis-aligned bounding boxes in X and Y directions. Can individually adjust horizontal and vertical expansion ranges, generally suitable for extracting elongated objects.

  • Enable reverse: Reverses extraction logic, default is disable, enabling can shrink extraction range.