It can be observed from Fig. 2 that the proposed color correction algorithm is composed of three modules, namely the selection of color samples for training, the selection of the optimized BP neural network by EMC, and the training of color correction models based on MEC-BP-Adaboost neural network