Depth estimation of multi-depth objects based on computational ghost imaging system

Abstract Computational ghost imaging (CGI) can estimate the depth of an object by evaluating the degree of defocus in the reconstructed images. However, this technology has not yet realized the depth estimation of multi-depth objects, because the defocus and in-focus phenomenon of different objects may be observed in a reconstructed image, which affects the performance of the evaluation functions. In this paper, we first analyze the formed images of multi-depth object and select the gradient domain as the image transformation space. The images are formed by the algorithm of compressed sensing based on TV norm which suppress the background noise and the effect of the defocused image. Furthermore, the deviation-based correlation (DBC) is chosen to evaluate the degree of defocus. Finally, in order to improve the efficiency, we propose the depth estimation strategy using the variable resolution speckles, which reduces the required depth slices by ∼ 42 % for detecting the object of interest in a given system. This research promotes the application of CGI in the field of depth-imaging.