For the purpose of enhancing the detecting stability and the model adaptability of egg freshness by non-destructive detection method, a sensor fusion was taken by the machine vision and electronic nose in the sensor level of characteristics while D—S evidence theory was chosen as the sensor information fusion method and BP artificial neural network as the specific modeling method. An improved method that could remedy for the deficiency of D—S evidence theory was discussed. Verification results showed that the basic probability assignment of uncertainty decreased to less than 0.01 by sensor fusion optimization. The problem of low detecting range in single sensor method has been well solved. Meanwhile, the egg freshness discriminating accuracy and stability has been improved compared with no sensor fusion situation. The average discriminating accuracy reached to 92.6%.