Linear Spatio-Temporal Scale-Space
SCALE-SPACE '97 Proceedings of the First International Conference on Scale-Space Theory in Computer Vision
Integral Invariants for Shape Matching
IEEE Transactions on Pattern Analysis and Machine Intelligence
Region matching with missing parts
Image and Vision Computing
Hi-index | 0.00 |
Our aim is to compare two mammograms (left-right, temporal) in an unsupervised manner. To this end, we propose a novel region matching algorithm (RMA) for mammograms based upon the non-emergence and non-enhancement of maxima and the causality principle of integral invariant scale space (in a limited sense). The algorithm has several advantages over commonly used methods for comparing segmented regions as shapes. First, it gives improved key-points alignment for optimal shape correspondence. Second, it identifies new growths and complete/partial occlusion in corresponding regions by dividing the segmented region into sub-regions based upon the extrema that persist over all scales. Third, the algorithm does not depend upon the spatial locations of mammographic features and eliminates the need for registration to identify salient changes over time. Finally, the algorithm is fast to compute and requires no human intervention.