Miller Research on Graph Connectivity Analysis

Graph Convergence Through Motion

This research shows that a moving object can be extracted by treating image motion as a connectivity problem on a weighted Region Adjacency Graph. The frame is first divided into homogeneous regions; every region becomes a node and every shared boundary becomes an edge.

The central observation is that the moving object does not have to be fully visible as motion in every single instant. Across the time axis, repeated comparisons accumulate evidence until the true object becomes a stable connected graph component.

RAG = regions as nodes Temporal edge evidence GCP weighting NCA contraction Boundary stabilization

Proof that the graph converges

The convergence follows from the difference between persistent object evidence and temporary occlusion evidence. For object edges, high change probability remains stable across multiple comparisons; for occluded edges, the evidence appears only as a temporary peak and then declines. The GCP mapping preserves the first behavior and suppresses the second.

Local evidence LCPet,t+i = |e|-1Σpxlt,t+i measures boundary change on each graph edge.
Object edge LCPet,t+i ≈ high for many iterations, because the object's original boundary keeps producing change evidence.
Occlusion edge LCPet,t+k = GM only near a temporary coverage event; therefore its distribution is unstable.
GCP filter GCPei = μ(LCP) − σ(GM − LCP) rewards persistent evidence and penalizes temporary peaks.
Convergence D(∂Obji, ∂Obji−1) → ε, so the connected node set stabilizes into an accurate object representation.