Fig. 27: Flowchart of the deep-learning-based end-to-end disparity prediction method and the predicted disparity map result. | Light: Science & Applications

Fig. 27: Flowchart of the deep-learning-based end-to-end disparity prediction method and the predicted disparity map result.

From: Deep learning in optical metrology: a review

Fig. 27

a The Flowchart of the deep-learning-based end-to-end disparity prediction method: stereo images are fed into the constructed GC-Net to directly output disparity images of two perspectives. b The left input. c The disparity predicted by deep learning. d Ground truth. a–d ©(2021) IEEE. Adapted, with permission, from ref. 389

Back to article page