supported by National Natural Science Foundation of China(Grant Nos.61402489,61972435,61972435,61602499);Natural Science Foundation of Guangdong Province(Grant No.2019A1515011271);Fundamental Research Funds for the Central Universities(Grant No.18lgzd06);Shenzhen Technology and Innovation Committee(Grant No.201908073000399)。
Deep neural networks have shown great success in stereo matching in recent years.On the KITTI datasets,most top performing methods are based on neural networks.However,on the Middlebury datasets,these methods usually ...
supported by National Natural Science Foundation of China(Grant No.61272326);Startup Foundation for Introducing Talent of NUIST(Grant No.2243141601013)
Although the stereo matching problem has been extensively studied during the past decades, automatically computing a dense 3D reconstruction from several multiple views is still a difficult task owing to the problems ...