The VLSI implementation of a high-resolution depth-sensing SoC based on active structured light

Spatial Localization of EEG Electrodes in a TOF+CCD Camera System

“…Recently, there has been a great deal of interest in the development and applications of time-of-flight (TOF) depth cameras. In 2015, Yao et al presented the full very large-scale integration (VLSI) implementation of a new high-resolution depth-sensing system on a chip (SoC) based on active infrared structured light, which estimates the 3D scene depth by matching randomized speckle patterns (Yao et al, 2015). At the same year, Golbach et al presented a computer-vision system for seedling phenotyping that combines best of both approaches by utilizing TOF depth cameras (Golbach et al, 2016).…”

A High Spatial Resolution Depth Sensing Method Based on Binocular Structured Light

“…Hence, a consistency enhancement algorithm proposed in [20] is used to enhance the input speckle patterns to make it more discriminative so as to improve the matching accuracy. The algorithm combines grayscale transformation and histogram equalization, is suitable for hardware implementation, and can be described as $f^{*}\left(x,y\right)=\{\begin{array}{cc}\beta \times \left(f\left(x,y\right)-f\left(x,y\right)\right),& \mathrm{if} f\left(x,y\right)\ge \overline{f}\left(x,y\right)\\ 0,& \mathrm{if} f\left(x,y\right)\le \overline{f}\left(x,y\right)\end{array}$ where $\overline{f}\left(x,y\right)$ is the average gray value of the subset with the center pixel $\left(x,y\right)$ and $\beta =gra{y}_{ref}/\overline{f}\left(x,y\right)$ is a scale factor.…”

“…In [ 20 ], a consistency enhancement algorithm is proposed to make the intensity of the pattern formed at different distances as consistent as possible so as to ensure the matching accuracy in the disparity estimation process. However, the surface material of the objects also affects the projected pattern, so we have to select a larger block, compared with the smallest size in theory, to perform the block-matching step.…”

“…Moreover, the hardware implementation of depth sensors based on structured light remains a black box, to a large extent, in the literature. In [ 20 ], we have proposed a full very large scale integration (VLSI) implementation method to obtain a high-resolution and high-accuracy depth map based on randomized speckle patterns. Then it was found that the spatial resolution of the depth map has the potential to be further improved.…”

“…Our key contribution in this paper, beyond [ 20 ], is that an infrared camera is added at the right of the laser projector, then the binocular matching mode, performed between the captured left and right patterns, is employed to improve the spatial resolution in the X-Y direction. The mismatching and occlusion is unavoidable in the binocular matching process, so the monocular matching mode, performed between the left and the reference patterns, is also employed to revise the matching results.…”