The next-best-view is the sensor location which maximizes the sensed Surface during the three-dimensional object reconstruction task. Here, we present a dataset for Next-best-view planning for 3D object reconstruction. The idea of this dataset is to leverage the research on new machine learning based techniques. This dataset was build by exahustive computations. More details are find in our paper (under review): Miguel Mendoza, J. Irving Vasquez-Gomez, Hind Taud, Luis Enrique Sucar, Carolina Reta, Supervised Learning of the Next-Best-View for 3D Object Reconstruction, arXiv:1905.05833 Aknowledgements This dataset is based on the range images provided by Hinterstoisser et al [1] [1] Hinterstoisser et al. "Model based training, detection and pose estimation of texture-less 3d objects in heavily cluttered scenes" ACCV