Abstract：The activation function can increase the nonlinearity of the algorithm and improve the complexity of the algorithm in the super-resolution reconstruction algorithm. Using the ReLU activation function, the algorithm has the advantage of short training time, and in view of the problem that the negative value is zeroed through the ReLU activation function, resulting in the inactivation of some neurons, the inactivated part is re-added to the model through the rReLU function, and the method is named FSRCNN supplementary module. The results show that the peak signal-to-noise ratio of the supplementary module algorithm is 0.1dB higher than that of the original FSRCNN algorithm under the condition of magnification of 4, so the supplementary module can improve the performance of the model and enhance the extraction of information by the model.
He K M, Zhang X, Ren S, et al. Deep Residual Learning for Image Recognition ［C］//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, USA, 2016.
［11］
Wei Y, Gu S, Li Y, et al. Unsupervised Real-world Image Super Resolution via Domain-distance Aware Training ［C］//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, USA, 2021.
Zeng X Y, Lu H C, Zhang C Y. Super Resolution Image Restoration Algorithm: Based on Wavelet and Interpolation［C］//2020 3rd World Conference on Mechanical Engineering and Intelligent Manufacturing (WCMEIM). Shanghai, China, 2020, 95-98.
［3］
Yu J, Gao X, Tao D, et al. A Unified Learning Framework for Single Image Super-Resolution ［J］. IEEE Transactions on Neural Networks and Learning Systems, 2014, 25 (4): 780-792.
Shi W, Caballero J, Huszár F, et al. Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network［C］//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, USA, 2016.
［10］
Lim B, Son S and Kim H, et al. Enhanced Deep Residual Networks for Single Image Super-Resolution ［C］//2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Honolulu, USA, 2017.
［12］
Zhu Y, Zhou Z, Liao G, et al. Csrgan: Medical Image Super-Resolution Using A Generative Adversarial Network［C］//2020 IEEE 17th International Symposium on Biomedical Imaging Workshops (ISBI Workshops). Iowa City, USA, 2020.
Zhang L G, Jiang T X, Tian G J. Research on Unmanned Aerial Vehicle to Ground Vehicle Target Detection Algorithm Based on Multiscale Fusion Method ［J］. Acta Metrologica Sinica, 2021, 42 (11): 1436-1442.
Zhang J G, Ta Y Y, Jiang R J, et al. Improvement of the Richardson-Lucy Algorithm in Blurred Image Restoration. Acta Metrologica Sinica, 2020, 41 (2): 153-158.
Cheng S H, Lu J X, Zhang D F, et al. Wheel Hub Identification of Convolutional Neural Networks Based on Ring Feature［J］. Acta Metrologica Sinica, 2022, 43 (6): 730-736.
Zhu Q G, Dong H R, Zhang M Y. Humanoid Robot Motion Imitation Based on Human Posture Recognition ［J］. Acta Metrologica Sinica, 2021, 42 (9): 1136-1141.
［6］
Dong C, Loy C C, He K, et al. Image Super-Resolution Using Deep Convolutional Networks ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2016, 38 (2): 295-307.
［13］
Zhang Y, Tian Y, Kong Y, et al. Residual Dense Network for Image Restoration ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43 (7): 2480-2495.
Zhang S H, Yan X R, Sang Y. Single Image Rain Removal Method by Fusing Residual and Channel Attention Mechanism ［J］. Acta Metrologica Sinica, 2021, 42 (1): 20-28.
［15］
Li Y, Si H Q, Zong Y T, et al. Application of Neural Network Based on Real-Time Recursive Learning and Kalman Filter in Flight Data Identification ［J］. International Journal of Aeronautical and Space Sciences, 2021, 22 (6): 1383-1396.
Zhao L Y, Wu Y Q. Research progress of surface defect detection methods based on machine vision. Chinese Journal of Scientific Instrument, 2022, 43 (1): 198-219.
2023, Vol. 44 
