Boosting Object Recognition in Point Clouds by Saliency Detection

Abstract

Object recognition in 3D point clouds is a challenging task, mainly when time is an important factor to deal with, such as in industrial applications. Local descriptors are an amenable choice whenever the 6 DoF pose of recognized objects should also be estimated. However, the pipeline for this kind of descriptors is highly time-consuming. In this work, we propose an update to the traditional pipeline, by adding a preliminary filtering stage referred to as saliency boost. We perform tests on a standard object recognition benchmark by considering four keypoint detectors and four local descriptors, in order to compare time and recognition performance between the traditional pipeline and the boosted one. Results on time show that the boosted pipeline could turn out up to 5 times faster, with the recognition rate improving in most of the cases and exhibiting only a slight decrease in the others. These results suggest that the boosted pipeline can speed-up processing time substantially with limited impacts or even benefits in recognition accuracy.

Towards Real-time Object Recognition and Pose Estimation in Point Clouds

“…However, with a descriptor length of 1344 bins, it is not suitable for real-time featurematching. Another proposal that deals with color is PFHRGB (Rusu et al, 2008), which, despite being shorter (250 bins) than CSHOT, presents inefficient calculation time (Marcon et al, 2019).…”

“…We also evaluate the correctness of predictions, both of object presence and pose. To do so, we follow (Marcon et al, 2019) and employ the Intersection over Union metric (IoU), defined as:…”

“…We compared our proposal with the standard 3D object recognition and pose estimation pipeline (Aldoma et al, 2012b), and with a boosted version of such pipeline, proposed by (Marcon et al, 2019). To calculate precision-recall curves (PRC), we varied the threshold on the minimum geometrically consistent correspondences, starting from at least three, related to each object's best-suited partial view.…”

“…Deep learning models were implemented in PyTorch, and the pre-trained models extracted from torchvision. To implement traditional and boosted versions of object recognition and pose estimation pipelines, we use PCL 1.8.1, OpenCV 3.4.2, and the saliency detection of (Hou et al, 2017), following (Marcon et al, 2019).…”

“…Unfortunately, for large or dense point clouds, its runtime becomes a significant limitation in several of the example applications mentioned above (Vock et al, 2019). When we seek a 6Dof estimation pose, the real-time is a more challenging task (Marcon et al, 2019). In an extensive benchmark of full cloud object detection and pose estimation, (Hodan et al, 2018) reported runtime of a second per test target on average.…”