ARobust Stamp Detection Framework on Degraded Documents

In this paper, we present a novel stamp detection framework based on parameter estimation of connected edge features. Using robust basic shape detectors, the approach is effective for stamps with analytically shaped contours. For elliptic/circular stamps, it efficiently makes use of the orientation information from pairs of edge points to determine its center position and area without computing all the five parameters of ellipse. When linking connected edges, we introduce an effective template-based junction removal technique to specifically address the problem that stamps often spatially overlap with background contents. These give our approach significant advantage in terms of computation complexity over traditional Hough transform methods. Experimental results on real degraded documents demonstrated the robustness of this retrieval approach on large document database, which consists of both printed text and handwritten notes. This enables us to reliably retrieve documents from a specific source by detecting its official stamps with analytically-shape contours when only limited samples are available.