Single-pixel three-dimensional imaging based on 2D compressed sensing algorithm.

In recent years, non-scanning three-dimensional imaging technology has become increasingly widespread in the field of rapid detection and recognition. Due to the characteristics of integrating signal compression and sampling, compressed sensing only requires a small amount of measurement data to reconstruct the original signal, making itself highly suitable for fast three-dimensional imaging. Based on the proposed new two-dimensional compressed sensing algorithm V2DALM, combined with the time-of-flight measurement principle, this paper achieves single-pixel, non-scanning imaging of three-dimensional objects. The feasibility of algorithm V2DALM in 3D imaging is verified by the principle experimental results. Compared with the other two algorithms 2DPG-ED and TVAL3 in simulation, it can be found that under similar imaging quality, the reconstruction time of algorithm V2DALM is significantly shorter than that of TVAL3; under the same sampling rate, the noise resistance of algorithm V2DALM is significantly better than that of 2DPG-ED. Consequently, this method can be potentially applied to facilitate the rapid reconstruction of high-resolution depth maps of three-dimensional objects.