Point Cloud Generation From Multiple Angles of Voxel Grids

Abstract

The advancement of deep learning technology has been concentrating on deploying end-to-end solutions using high dimensional data, such as images. Recently, a number of methods have been proposed for reconstructing 3D objects using deep learning. One such technique is the method that involves recovering 3D objects as voxel grids using one or multiple images. However, there has been very little work directed towards the generation of 3D objects represented by a set of points, i.e. point cloud, from voxel grids which are ambiguous and coarse. The development of a deep learning model that generates point clouds with details from coarse voxel grids has numerous benefits to quality of life. For example, design professionals can use this model to generate detailed 3D point clouds using a sketched and coarse voxel to enable their creativities. This paper presents an algorithm to generate point clouds from voxel grids. The algorithm explicitly loads 3D objects into voxels without projection operations and associated information loss. To obtain a comprehensive understanding of the voxel grid, the grid is analyzed through various angles, as inspired by how humans observe 3D objects. The features from various angles are passed into a GRU layer to extract patterns across views, which will then be passed to a channel-wise convolutional layer and graph convolution to generated the predicted point cloud. The experimental result of the algorithm indicates that the algorithm is capable of generating high-quality point clouds by understanding the semantic features of the voxel grid. INDEX TERMS 3D object generation, neural networks, point clouds.

Simultaneous Localization and Mapping (SLAM) for Synthetic Aperture Radar (SAR) Processing in the Field of Autonomous Driving

“…In this case d mi describes the distance between the n r -th radar sensor at the global ( g ) position P g = P g x , P g y and the cell m i The final SAR image is processed exclusively in the xy-plane at the height of the radar sensors, given that the radar sensors, along with the vehicle's motion, cover a synthetic aperture limited to the xy-plane. According to (8), each cell of the map to be imaged is assigned a distance and thus also a complex-value received power according to (6). This complex pointer must be rotated around the phase Φ according to the backprojection algorithm (7) to obtain a constructive or destructive overlay.…”

“…The target information obtained from radar raw data can be used to create environmental maps, e.g. grid maps [6]. These offer the advantage of a cell-based representation of the environment, which makes them efficient to process.…”

Probabilistic SAR Processing for High-Resolution Mapping Using Millimeter-Wave Radar Sensors

“…The target information, determined from radar raw data, can be used with the help of different methods for environmental mapping. A fundamental differentiation is being made between point clouds and grid maps [5]. Grid maps offer the advantage of a cell-based representation of the environment, which allows efficient processing in applications based on them.…”

Autonomous Scene Exploration for Robotics: A Conditional Random View-Sampling and Evaluation Using a Voxel-Sorting Mechanism for Efficient Ray Casting

“…That being said, some other methods also use the same philosophy of dividing the space into a 3D grid, called voxel grids [10]. A voxel grid is generated when the scene is divided into several, recursively smaller, cubic volumes, usually called voxels (hence the name voxel grid).…”