Fast adaptive compressed 3D imaging based on wavelet trees and Hadamard matrix

Aiming at the problem of slow 3D imaging speed in photon counting system, a method of fast adaptive compressed 3D imaging method based on wavelet trees and Hadamard matrix was proposed. The sampling efficiency was improved by modulating projective patterns with Hadamard matrix, the scene was illuminated with modulated short-pulsed structured light, and the echoes were collected by the single-pixel photon counting detector. The sample area was selected by analyzing the wavelet trees of coarser images. With the patterns projected from the sample area modulated by Hadamard matrix, image details could be sampled. After the multistage sampling, the high-resolution image could be recovered with Hadamard inverse transform. The experimental results indicate that a 3D image at resolution up to 512×512 pixel can be acquired and retrieved with practical time as low as 41 s.