Rethinking the sGLOH Descriptor

Abstract

Abstract-sGLOH (shifting GLOH) is a histogram-based keypoint descriptor that can be associated to multiple quantized rotations of the keypoint patch without any recomputation. This property can be exploited to define the best distance between two descriptor vectors, thus avoiding computing the dominant orientation. In addition, sGLOH can reject incongruous correspondences by adding a global constraint on the rotations either as an a priori knowledge or based on the data. This paper thoroughly reconsiders sGLOH and improves it in terms of robustness, speed and descriptor dimension. The revised sGLOH embeds more quantized rotations, thus yielding more correct matches. A novel fast matching scheme is also designed, which significantly reduces both computation time and memory usage. In addition, a new binarization technique based on comparisons inside each descriptor histogram is defined, yielding a more compact, faster, yet robust alternative. Results on an exhaustive comparative experimental evaluation show that the revised sGLOH descriptor incorporating the above ideas and combining them according to task requirements, improves in most cases the state of the art in both image matching and object recognition.

Which Is Which? Evaluation of Local Descriptors for Image Matching in Real-World Scenarios

“…Such matrix reflects the kind of distance employed by the descriptor, but can also allow one to exploit descriptor distance statistics inside images, as done in [4]. Finally, the distance table was employed to extract the best matches according to their distance in a greedy way, so as to avoid that two matches share a common keypoint.…”

“…Seven local image descriptors were submitted to WISW. These include SOS-Net [35], still unpublished at contest time, the recent HardNet A [29], obtained by training HardNet [24] on AMOS [29] and other datasets, RalNet Shuffle using the RalNet architecture [38] and additionally cropping and shuffling patches at training time, and RsGLOH2, "square rooting" sGLOH2 [4] according to RootSIFT [1]. Two variants of HardNet A , exploiting the deep networks described in [25] either for custom orientation assignment or to accommodate patches before extracting the descriptor, were also submitted as OriNet+HardNet A and AffNet+HardNet A , respectively.…”

Prnu Pattern Alignment for Images and Videos Based on Scene Content

“…For the registration, corner-like keypoints extracted with the HarrisZ detector [11] are matched with the recent SIFTlike sGLOH2 local image descriptor [12], and the initial geometric transformation parameters are estimated using RANSAC [9]. As shown in the bottom row of Fig.…”

FISH: Face intensity-shape histogram representation for automatic face splicing detection

“…In particular, instead of computing analytically the lighting coefficients, we build histograms relating surface normals with their intensity values, by statistically modelling the interaction map between light and the surface. The resulting descriptor design is inspired by histogram-based keypoint descriptors [13] employed in robust image matching. Indeed, the histograms associated to different faces are stable and robust to shape variations, and can be successfully used to indirectly measure lighting inconsistencies between spliced and pristine faces.…”

“…• Instead of a complex and partially incomplete ideal model characterizing the interaction of light with faces, we propose to employ histogram-based features. Histograms have proved to be very effective in many computer vision tasks [13] and, to the best of our knowledge, were never employed for face splicing detection;…”