Issue No. 10 - Oct. (2012 vol. 34)
Bin Fan , Chinese Academy of Sciences, Beijing
Fuchao Wu , Chinese Academy of Sciences, Beijing
Zhanyi Hu , Chinese Academy of Sciences, Beijing
This paper proposes a novel method for interest region description which pools local features based on their intensity orders in multiple support regions. Pooling by intensity orders is not only invariant to rotation and monotonic intensity changes, but also encodes ordinal information into a descriptor. Two kinds of local features are used in this paper, one based on gradients and the other on intensities; hence, two descriptors are obtained: the Multisupport Region Order-Based Gradient Histogram (MROGH) and the Multisupport Region Rotation and Intensity Monotonic Invariant Descriptor (MRRID). Thanks to the intensity order pooling scheme, the two descriptors are rotation invariant without estimating a reference orientation, which appears to be a major error source for most of the existing methods, such as Scale Invariant Feature Transform (SIFT), SURF, and DAISY. Promising experimental results on image matching and object recognition demonstrate the effectiveness of the proposed descriptors compared to state-of-the-art descriptors.
Image matching, Object recognition, Estimation error, Detectors, Robustness, Feature extraction, SIFT., Local image descriptor, rotation invariance, monotonic intensity invariance, image matching, intensity orders
Z. Hu, F. Wu and B. Fan, "Rotationally Invariant Descriptors Using Intensity Order Pooling," in IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 34, no. , pp. 2031-2045, 2012.