Image Segmentation from Shadow-Hints using Minimum Spanning Trees

TL;DR

This work proposes a novel image segmentation method, achieving similar segmentation quality but without training, that requires an image sequence with a static camera and a single light source at varying positions, as used in for photometric stereo, for example.

Abstract

Image segmentation in RGB space is a notoriously difficult task where state-of-the-art methods are trained on thousands or even millions of annotated images. While the performance is impressive, it is still not perfect. We propose a novel image segmentation method, achieving similar segmentation quality but without training. Instead, we require an image sequence with a static camera and a single light source at varying positions, as used in for photometric stereo, for example.

Figures (1)

• Figure 1: From Top to Bottom: RGB input used to generate segmentations with FH04 and SAM23 as well as our segmentation from shadow-hints. For our method, detected outline points are overlaid to visualize where completions occur.