Aiming at problem of low detection accuracy of vaccine medium

first propose using machine vision

multi-sensor information fusion technology of infrared temperature and illumination to detect vaccine medium. Through BP neural network

information collected by three kinds of sensors is fused in feature level

in process of model building

numbers of principal components factor of each message are optimized. From experimental results

detection recognition rate of model on prediction set is 97 %

test results show that detecting vaccine medium through multi-information fusion method is feasible

accuracy and stability of test results are improved significantly relative to individual model.