Invariant Hough Random Ferns for Object Detection and Tracking

Abstract

This paper introduces an invariant Hough random ferns (IHRF) incorporating rotation and scale invariance into the local feature description, random ferns classifier training, and Hough voting stages. It is especially suited for object detection under changes in object appearance and scale, partial occlusions, and pose variations. The efficacy of this approach is validated through experiments on a large set of challenging benchmark datasets, and the results demonstrate that the proposed method outperforms state-of-the-art conventional methods such as bounding-box-based and part-based methods. Additionally, we also propose an efficient clustering scheme based on the local patches’ appearance and their geometric relations that can provide pixel-accurate, top-down segmentations from IHRF back-projections. This refined segmentation can be used to improve the quality of online object tracking because it avoids the drifting problem. Thus, an online tracking framework based on IHRF, which is trained and updated in each frame to distinguish and segment the object from the background, is established. Finally, the experimental results on both object segmentation and long-term object tracking show that this method yields accurate and robust tracking performance in a variety of complex scenarios, especially in cases of severe occlusions and nonrigid deformations.

A neural implementation of the Hough transform and the advantages of explaining away

“…The current work has been concerned with improving the Hough voting process. Other recent work on the HT has been concerned with improving the identification of voting elements (e.g., Gall and Lem-pitsky, 2009;Gall et al, 2011;Lin et al, 2014;Maji and Malik, 2009;Okada, 2009;Razavi et al, 2011). This previous work is entirely complementary to the current work, and it should therefore be possible to combine the proposed voting process with these improved methods of voting element identification to further boost the performance of the HT.…”

“…Errors in the estimate of the orientation parameter (θ) are shown in the two graphs on the left, and errors in the estimate of the radius parameter (ρ) are shown in the two graphs on the right. ISM (Leibe et al, 2008) 9 -ISM+MDL verification (Leibe et al, 2008) 2.5 5 Hough Forest (Gall and Lempitsky, 2009;Gall et al, 2011) 1.5 2.4 Discriminative HT (Okada, 2009) 1.5 -IHRF (Lin et al, 2014) 0 1.3 PRISM (Lehmann et al, 2011) -2.2 ESS (Lampert et al, 2008) 1.5 1.4 sliding window HMAX+verification (Mutch and Lowe, 2006) 0.06 9.4 chains model (Karlinsky et al, 2010) 0.5 - car is in the centre of the image. For the second image, the standard method of voting results in four peaks.…”

Invariant Hough random ferns for RGB-D-based object detection

“…12 So, the Fern size of 10 and 13 features used for each Fern have proved to be a good compromise in our recent experiments. 11 The RGB-D Object Dataset 17 is a large dataset of 300 common household objects, eight of which were used for training and detection, as shown in Fig. 4.…”

“…However, it has been reported that Hough voting efficiency during object categorization can be improved using a highly efficient classifier. 10 With regard to invariant Hough random ferns (IHRF), 11 this paper applies a random ferns classifier (RFC) 12 to a Hough transform to improve search speed and reduce the need for a large storage space for data. Furthermore, the Hough voting is performed in rotation-invariant Hough space, since each support point shares a stable polar angle and scalable displacement related to the center of the relevant object.…”