0.003。Computer vision analysis and impact excitation frequency analysis were used to construct the duck egg’s intensity model. By the analysis in the CV (coefficient of variation) of the duck egg’s impact excitation frequency and the analysis of the crack egg’s intensity, the threshold value (λ=0.02) which can be applied to distinguish the duck egg intensity at the level of 1MPa (confidence at the level of 95%) was obtained. Then the characteristic variables were entered into the model. The backward elimination procedure was used to perform multi-dimensional linear regression in order to establish the forecast model of the non-crack duck eggs’ shell intensity. Combined with the threshold value which had been obtained, the system model of duck egg’s shell intensity based on the multi-sensor fusion can be obtained. The confirmation experiment indicates that the model had good adaptability on duckeggs with different color and mass, the model shows high accuracy (the mean of forecast error is 3.1%) as well as good stability (the standard deviation of forecast error is 0.003). 