Abstract:Binary descriptors are becoming widely used in computer vision field because of their high matching efficiency and low memory requirements. Since conventional approaches, which first compute a floating-point descriptor then binarize it, are computationally expensive, some recent efforts have focused on directly computing binary descriptors from local image patches. Although these binary descriptors enable a significant speedup in processing time, their performances usually drop a lot due to orientation estimation errors and limited description abilities. To address these issues, we propose a novel binary descriptor based on the ordinal and spatial information of regional invariants (OSRIs) over a rotation invariant sampling pattern. Our main contributions are twofold: 1) each bit in OSRI is computed based on difference tests of regional invariants over pairwise sampling-regions instead of difference tests of pixel intensities commonly used in existing binary descriptors, which can significantly enhance the discriminative ability and 2) rotation and illumination changes are handled well by ordering pixels according to their intensities and gradient orientations, meanwhile, which is also more reliable than those methods that resort to a reference orientation for rotation invariance. Besides, a statistical analysis of discriminative abilities of different parts in the descriptor is conducted to design a cascade filter which can reject nonmatching descriptors at early stages by comparing just a small portion of the whole descriptor, further reducing the matching time. Extensive experiments on four challenging data sets (Oxford, 53 Objects, ZuBuD, and Kentucky) show that OSRI significantly outperforms two state-of-the-art binary descriptors (FREAK and ORB). The matching performance of OSRI with only 512 bits is also better than the well-known floating-point descriptor SIFT (4K bits) and is comparable with the state-of-the-art floating-point descriptor MROGH (6K bits), while it is two orders of magnitude faster to match than SIFT and MROGH.

A Rotationally Invariant Descriptor Based on Mixed Intensity Feature Histograms

A Robust Method for Constructing Rotational Invariant Descriptors

Description of Interest Regions Based on Combined Local Binary Patterns

Robust Method for Interest Region Description Based on Local Intensity Binary Pattern

An Efficient and Robust Rotation Invariant Descriptor

RISAS: A Novel Rotation, Illumination, Scale Invariant Appearance and Shape Feature.

Robust Local Feature Descriptor for Multisource Remote Sensing Image Registration

RI-LPOH: Rotation-Invariant Local Phase Orientation Histogram for Multi-Modal Image Matching

LOIND: an Illumination and Scale Invariant RGB-D Descriptor

Srifd: A Shift Rotation Invariant Feature Descriptor for Multi-Sensor Image Matching

An Advanced Rotation Invariant Descriptor for SAR Image Registration.

A Novel Local Pettern Based Self-Similarity Descriptor for Multisource Remote Sensing Image Registration.

OSRI: A Rotationally Invariant Binary Descriptor

Local directional derivative pattern for rotation invariant texture classification

A Rotation Invariant Descriptor for Robust Video Copy Detection

Reflection and Rotation Invariant Uniform Patterns for Texture Classification

Learning Rotation-Invariant Local Binary Descriptor.

Efficient and distinctive binary descriptor for rotated circular image recognition

MFD: Mutual feature description for image matching

An Effective Local Invariant Descriptor Combining Luminance and Color Information

Local Invariant Descriptor for Image Matching