This paper presents a new Bayesian model and algorithm used for depth andintensity profiling using full waveforms from the time-correlated single photoncounting (TCSPC) measurement in the limit of very low photon counts. The modelproposed represents each Lidar waveform as a combination of a known impulseresponse, weighted by the target intensity, and an unknown constant background,corrupted by Poisson noise. Prior knowledge about the problem is embedded in ahierarchical model that describes the dependence structure between the modelparameters and their constraints. In particular, a gamma Markov random field(MRF) is used to model the joint distribution of the target intensity, and asecond MRF is used to model the distribution of the target depth, which areboth expected to exhibit significant spatial correlations. An adaptive Markovchain Monte Carlo algorithm is then proposed to compute the Bayesian estimatesof interest and perform Bayesian inference. This algorithm is equipped with astochastic optimization adaptation mechanism that automatically adjusts theparameters of the MRFs by maximum marginal likelihood estimation. Finally, thebenefits of the proposed methodology are demonstrated through a serie ofexperiments using real data.
展开▼
