In the process of studying the Tang Sancai figurines, some images will be degraded due to optical system, motion, atmospheric turbulence, etc., so the images need to be restored. With the best restoration method, the restored image can meet the requirements. In fact, the purpose of image restoration is to process the degraded image to make the restored image closer to the original image. This paper conducts a comparative experiment on the classical image reconstruction methods, taking the images of Tang Sancai figurines as the experimental objects. The results show that the image reconstruction quality of the least squares method is the best among the methods selected for the experiment in this paper, and the SSIM and PSNR index values of the reconstructed image A are 0.9612 and 31.7612, respectively; in the performance comparison of GAN, GA-GAN, and Dense-GAN models, the image reconstruction algorithm based on the GA-GAN model has the best performance. Among the ten images of Tang Sancai figurines used in the experiment, the highest SIMM value is 0.99, and the highest PSNR value is 27.9345.

Shengwei Qiu. Image Reconstruction of Tang Sancai Figurines Based on Artificial Intelligence Image Extraction Technology Based on Ration. International Journal of Multimedia Computing (2023), Vol. 4, Issue 1: 37-53. https://doi.org/10.38007/IJMC.2023.040103.

[1] Zhang J, Zhang Y. Design and Implementation of Edge Extraction Algorithm for Digital Image. International Core Journal of Engineering. (2019) 5(9): 105-115.

[2] Hu S. Research on Data Acquisition Algorithms Based on Image Processing and Artificial Intelligence. International Journal of Pattern Recognition and Artificial Intelligence. (2020) 34(06): 1-13. https://doi.org/10.1142/S0218001420540166

[3] Ohanian T. How Artificial Intelligence and Machine Learning May Eventually Change Content Creation Methodologies. SMPTE Motion Imaging Journal. (2019) 128(1): 33-40. https://doi.org/10.5594/JMI.2018.2876781 

[4] Bhargavi V R, Rajesh V. Computer Aided Bright Lesion Classification in Fundus Image Based on Feature Extraction. International Journal of Pattern Recognition and Artificial Intelligence. (2018) 32(11): 1850034.1-1850034.15. https://doi.org/10.1142/S0218001418500349

[5] Wang X, Xie Q, Ma T, et al. Feature Extraction Based on Dimension Reduction and Clustering for Maize Leaf Spot Images. International Journal of Pattern Recognition & Artificial Intelligence. (2018) 32(12): 1854029.1-1854029.15. https://doi.org/10.1142/S0218001418540290

[6] Gupta H, Jin K H, Nguyen H Q, et al. CNN-Based Projected Gradient Descent for Consistent Image Reconstruction. IEEE Transactions on Medical Imaging. (2017) 37(6): 1-1. https://doi.org/10.1109/TMI.2018.2832656

[7] Oliveira E, Gama J, Vale Z, et al. [Lecture Notes in Computer Science] Progress in Artificial Intelligence Volume 10423 || Image Matching Algorithm Based on Hashes Extraction. (2017) 10.1007/978-3-319-65340-2(Chapter 8): 87-94. https://doi.org/10.1007/978-3-319-65340-2_8

[8] Zhang X, Zhou J, Cherry S R, et al. Quantitative image reconstruction for total-body PET imaging using the 2-meter long EXPLORER scanner. Physics in Medicine and Biology. (2017) 62(6): 2465-2485. https://doi.org/10.1088/1361-6560/aa5e46

[9] Yunjie Yang, Jiabin, et al. A Miniature Electrical Impedance Tomography Sensor and 3-D Image Reconstruction for Cell Imaging. IEEE Sensors Journal. (2017) 17(2): 514-523. https://doi.org/10.1109/JSEN.2016.2631263

[10] Matthews T, Wang K, Li C, et al. Regularized Dual Averaging Image Reconstruction for Full-Wave Ultrasound Computed Tomography. IEEE Transactions on Ultrasonics Ferroelectrics & Frequency Control. (2017) 64(5): 811-825. https://doi.org/10.1109/TUFFC.2017.2682061

[11] Yu H, Shan X, Wang S, et al. Achieving High Spatial Resolution Surface Plasmon Resonance Microscopy with Image Reconstruction. Analytical Chemistry. (2017) 89(5): 2704-2707. https://doi.org/10.1021/acs.analchem.6b05049

[12] Ramon A J, Yang Y, Pretorius P H, et al. Investigation of dose reduction in cardiac perfusion SPECT via optimization and choice of the image reconstruction strategy. Journal of Nuclear Cardiology. (2017) 25(6): 1-12. https://doi.org/10.1007/s12350-017-0920-1

[13] Ozan, ktem, Chong, et al. Shape-based image reconstruction using linearized deformations. Inverse Problems. (2017) 33(3): 1-1. https://doi.org/10.1088/1361-6420/aa55af

[14] Ellis S, Reader A J. Simultaneous maximum a posteriori longitudinal PET image reconstruction. Physics in Medicine and Biology. (2017) 62(17): 6963-6979. https://doi.org/10.1088/1361-6560/aa7b49

[15] Yin J, Yu W, Jia X. Hybrid Norm Pursuit Method for Hyperspectral Image Reconstruction. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. (2017) 9(9): 1-9. https://doi.org/10.1109/JSTARS.2016.2535467

[16] Zhang H, Wang L, Duan Y, et al. Euler's Elastica Strategy for Limited-Angle Computed Tomography Image Reconstruction. IEEE Transactions on Nuclear Science. (2017) 64(8): 2395-2405. https://doi.org/10.1109/TNS.2017.2717864

[17] Lin Y, Shou K, Huang L. The initial study of LLS-based binocular stereo-vision system on underwater 3D image reconstruction in the laboratory. Journal of Marine Science & Technology. (2017) 22(3): 513-532. https://doi.org/10.1007/s00773-017-0432-3

[18] Chen H Y, Ng L S, Chang C S, et al. Pursuing Mirror Image Reconstruction in Unilateral Microtia: Customizing Auricular Framework by Application of Three-Dimensional Imaging and Three-Dimensional Printing. Plastic & Reconstructive Surgery. (2017) 139(6): 1433-1443. https://doi.org/10.1097/PRS.0000000000003374

[19] Guo Y, Li Z, Wang C, et al. Image reconstruction model for the exterior problem of computed tomography based on weighted directional total variation. Applied Mathematical Modelling. (2017) 52(dec.): 358-377. https://doi.org/10.1016/j.apm.2017.07.057

[20] Balakin D A, Belinsky A V, Chirkin A S. Improvement of the optical image reconstruction based on multiplexed quantum ghost images. Journal of Experimental & Theoretical Physics. (2017) 125(2): 210-222. https://doi.org/10.1134/S1063776117070147