An End-to-End Deep Learning Network for 3D Object Detection From RGB-D Data Based on Hough Voting

Abstract

Existing outdoor three-dimensional (3D) object detection algorithms mainly use a single type of sensor, for example, only using a monocular camera or radar point cloud. However, camera sensors are affected by light and lose depth information. When scanning a distant object or an occluded object, the data collected by the short-range radar point cloud sensor are very sparse, which affects the detection algorithm. To address the above challenges, we design a deep learning network that can combine the texture information of two-dimensional (2D) data and the geometric information of 3D data for object detection. To solve the problem of a single sensor, we use a reverse mapping layer and an aggregation layer to combine the texture information of RGB data with the geometric information of point cloud data and design a maximum pooling layer to deal with the input of multi-view cameras. In addition, to solve the defects of the 3D object detection algorithm based on the region proposal network (RPN) method, we use the Hough voting algorithm implemented by a deep neural network to suggest objects. Experimental results show that our algorithm has a 1.06% decrease in average precision (AP) compared to PointRCNN in easy car object detection, but our algorithm requires 37.7% less time to calculate than PointRCNN under the same hardware environment. Moreover, our algorithm improves the AP by 1.14% compared to PointRCNN in hard car object detection.

Hybrid features for object detection in RGB-D scenes

“…The target-domain detector improves the one-class universal detector by mining box-level pseudo ground facts in each iteration [20]. Yan et al [21] designed a deep learning network for object detection based on merging the geometric data (3D) and texture data of two-dimensional (2D). To solve the issue of one sensor, they used an inverse mapping level and a gathering level to merge the one or more input of RGB datum with the geometric input of point cloud data and designed a top gathering layer to transact with the data of multiple vision cameras.…”

“…Deductively, the type of dataset differs from one research to another based on the purpose of the research, whether to improve results or to a specific application, for example in [24] used object detection for surveillance purposes, they used specific objects classification in airports, such as people, bags, trolleys. While in [21], [25], [26], the authors studied 3D object detection using RGB-D data scenes in outdoor [27] and indoor [28]- [30], or both. some of selected related studies will be described in details in the following.…”

A Novel Vehicle Detection Framework Based on Parallel Vision

“…Recently, with the breakthrough progress of artificial intelligence technology, intelligent vehicles with the advanced driving assistance system (ADAS) are vigorously launched on the market [1,2]. The ADAS of the intelligent vehicle collects the surrounding data from the sensors like radars and cameras and then performs road object detection and so on.…”

SAL‐Net: Self‐Supervised Attribute Learning for Object Recognition and Segmentation

“…With the increasing demand for large-scale training sets for deep learning models [12][13][14][15][16][17][18][19][20][21][22], some researchers propose self-supervised learning methods [23][24][25][26], which train CNNs by automatic generation of labels based on the structure or characteristics of the image itself. We observe that each attribute of the object corresponds to a particular part of the object region.…”