Low Power Depth Estimation of Rigid Objects for Time-of-Flight Imaging

James Noraky,Vivienne Sze

DOI: https://doi.org/10.48550/arXiv.1810.01930

2019-03-26

Abstract:Depth sensing is useful in a variety of applications that range from augmented reality to robotics. Time-of-flight (TOF) cameras are appealing because they obtain dense depth measurements with minimal latency. However, for many battery-powered devices, the illumination source of a TOF camera is power hungry and can limit the battery life of the device. To address this issue, we present an algorithm that lowers the power for depth sensing by reducing the usage of the TOF camera and estimating depth maps using concurrently collected images. Our technique also adaptively controls the TOF camera and enables it when an accurate depth map cannot be estimated. To ensure that the overall system power for depth sensing is reduced, we design our algorithm to run on a low power embedded platform, where it outputs 640x480 depth maps at 30 frames per second. We evaluate our approach on several RGB-D datasets, where it produces depth maps with an overall mean relative error of 0.96% and reduces the usage of the TOF camera by 85%. When used with commercial TOF cameras, we estimate that our algorithm can lower the total power for depth sensing by up to 73%.

Image and Video Processing

What problem does this paper attempt to address?

The problem that this paper attempts to solve is that when using ToF (Time - of - Flight) cameras for depth perception on mobile devices, the battery life of the device is limited due to the high power consumption of its illumination source. Specifically, the paper proposes an algorithm to reduce the power consumption of depth perception by reducing the usage frequency of ToF cameras and using simultaneously - captured images to estimate the depth map. In addition, this algorithm can also adaptively control the ToF camera and enable it when the depth map cannot be accurately estimated. To ensure that the overall system reduces power consumption in depth perception, the algorithm is designed to run on a low - power embedded platform and output a 640×480 depth map at a rate of 30 frames per second.

The main contributions of the paper are:
1. Proposing a method to reduce the usage rate of ToF cameras by estimating 3D motion in the scene to update the previously measured depth map.
2. Reducing the amount of computation required for 3D motion estimation per pixel by estimating the inter - frame pose, and determining the optical flow using a block - matching heuristic method based on a sparse, uniformly - distributed grid.
3. Developing a mechanism to detect when the depth map cannot be accurately estimated and adaptively enabling the ToF camera in such cases.

The application of these techniques can significantly reduce the total power consumption of depth perception, especially when combined with commercial ToF cameras, it is estimated that up to 73% of the total power consumption can be reduced.