Fig. 1. Laser scanning projection two-axis galvanometer system

Fig. 2. Error of pillow

Fig. 3. The flow chart of BP neural network optimized by PSO

Fig. 4. The training flow chart of PSO-BP neural network

Fig. 5. Data acquisition through a binocular vision system

Fig. 6. The projection surface used in the experiment

Fig. 7. Curved surface diagram of error compensation

Fig. 8. The prediction accuracy of BP neural network with different number of hidden layers

Fig. 9. The training structure of BP neural network

Fig. 10. Variation curve of fitness value

Fig. 11. The fitting of PSO-BP neural network

Fig. 12. Structure of Elman neural neural network

Fig. 13. Flow chart of Elman neural network

Fig. 14. The prediction accuracy of Elman neural network with different number of hidden layers

Fig. 15. Fitting of Elman neural network

Fig. 16. Comparison of prediction accuracy between PSO-BP and Elman neural network