A deep convolutional neural network for continuous zoom with dual cameras
Hsueh-I Chen, Claire Chen, Yung‐Yu Chuang · 2017
This paper proposes a convolutional neural network for synthesizing continuous zoom from a pair of images captured by two nearby cameras with two fixed focal lengths. The network is composed of two sub-nets. The first one estimates the disparity map between input images so that they can be aligned. The second sub-net takes the aligned image pair for boosting the image resolution for simulating continuous zoom. We have built a prototype camera system and trained a network for the task. Experiments show that the proposed method can fuse the two captured images effectively and enhance the image resolution.