A series of hydrotalcite materials are prepared using hydrothermal synthesis method, and their spectral near-infrared water absorption peaks are tested using a visible near-infrared spectrometer to simulate the near-infrared water peaks in jungle camouflage coating materials. The influence of experimental conditions on the size of water peaks is explored. A PSO-RBF predictive model is constructed using partial serpentine spectroscopic data and verifies the reliability of the model. The validation results indicate that the predictive accuracy of model exceeds 90%, demonstrating excellent practical capability. Employing the predictive model to forecast the development of serpentine facilitates the determination of optimal preparation conditions, thereby reducing experimental time and further enhancing the water absorption peak of traditionally synthesized serpentine products for camouflage applications.