Hybrid reconstruction algorithm for computed tomography based on diagonal total variation

Deng, Luzhen; He, Peng; Jiang, Shanghai; Chen, Mian-Yi; Wei, Biao; Feng, Peng

doi:10.1007/s41365-018-0376-2

Cited by 2 publications

(2 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Instead of relying only on MSE-based loss, ISRGAN is trained by the combination of of VGG based perceptual loss and adversarial loss. To further improve the quality of super-resolution images, inspired by adaptive total variation [24], [25] and diagonal total variation [26], [27] model to take full advantage of directional information of edges and textures, a novel strategy is introduced to improve the performance of super-resolution based on GAN coupled with the total-variation based model. Furthermore, a new adaptive model based on diagonal total variation is proposed to keep texture details.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Adaptive Diagonal Total-Variation Generative Adversarial Network for Super-Resolution Imaging

et al. 2020

View full text Add to dashboard Cite

To address problems that the loss function does not correlate well with perceptual vision in super-resolution methods based on the convolutional neural network(CNN), a novel model called the ADTV-SRGAN is designed based on the adaptive diagonal total-variation generative adversarial network. Combined with global perception and the local structure adaptive method, spatial loss based on the diagonal variation model is proposed to make the loss function can be adjusted according to the spatial features. Pixel loss and characteristic loss are in combination with the spatial loss for the fusing optimization of the total loss function such that high-frequency details of the images are maintained to improve their quality. The results of experiment show that the proposed method can obtain competitive results in objective evaluations. In subjective assessment, images reconstructed by it are clear, delicate, and natural, and it preserved edgeand texture-related details.

show abstract

Section: Related Workmentioning

confidence: 99%

“…Spatial loss is based on the total-variation model [32], [33] to reconstruct an image that maintains features of the edges and texture to the greatest extent. The total-variation model is a classic model in image restoration that exhibits good edge retention characteristics [24]- [27]. Details of the image can be preserved while smoothing.…”

Section: Spatial Lossmentioning

confidence: 99%