This paper proposes a method to analyze the observed images of fluorescent images influenced by mutual illumination and estimate the spectral components. We suppose a general case where the entire surfaces of fluorescent objects have mutual illumination effects. First we model mutual illumination between the two objects. It is shown that the spectral composition is summarized with four components of (1) diffuse reflection, (2) diffuse-diffuse interreflection, (3) fluorescent self-luminescence, and (4) interreflection by mutual fluorescent illumination. Each component has two unknown factors of the spectral functions depending on wavelength and the weighting factors depending on pixel location. Second, an iterative algorithm is developed to solve this nonlinear estimation problem. Moreover, aiming a general solution which is independent of the initial conditions, we adopt a stabilization index to enforce the spectral smoothness and the spatial smoothness. Finally, the feasibility of the proposed method is shown using the spectral images of two adjacent fluorescent objects captured by a spectral imaging system in the visible range.