A Point Cloud Contour Extraction Method based on Plane Segmentation

This work extracts contour features from object point clouds through a two-stage pipeline. A PCA-guided RANSAC step segments the dominant planes; the segmented planar points are then projected to 2D, where gradient analysis recovers the surface contours. The method is aimed at industrial inspection, where the contours of interest are often incomplete or weakly defined.

This work appears in Frontiers in Computing and Intelligent Systems, 7(2):21-25.

Problem setting

A two-stage pipeline extracts contour features from object point clouds. First, the Random Sample Consensus (RANSAC) plane fit is guided by Principal Component Analysis (PCA) to segment the dominant planar regions. The segmented planar points are then projected to 2D, and contour features are extracted from the projection using gradients.

The figures below collect representative visual evidence from Frontiers in Computing and Intelligent Systems, 7(2):21-25.

Method and visual evidence

The visuals show PCA-guided RANSAC plane segmentation, 2D projection of the segmented plane, and gradient-based contour extraction examples.

Method overview.

Representation and setup.

Experimental evidence.

Result comparison.

Additional visual result.

Results and impact

The evaluation reported in Frontiers in Computing and Intelligent Systems, 7(2):21-25 is summarized through the figures above.