A global blind deconvolution estimation of fluorescence lifetime imaging microscopy measurements is proposed by using a multi-exponential model, constrained optimization, and alternating least-squares. A comprehensive validation is presented in two stages: fluorescence dyes and oral tissue samples.