Camvid Dataset is a driving labeled video dataset, which contains 700 pixel-level semantic segmentation and image segmentation. In order to ensure the accuracy of the data, each image has been checked and confirmed multiple times to ensure the accuracy of the data. This dataset addresses the need for experimental data and quantitatively evaluates emerging algorithms, where driving scenes increase the number and heterogeneity of observed objects. It provides more than 10 minutes of high-quality 30Hz footage and the corresponding semantically labeled images are 1Hz, and some are 15Hz.

This dataset was released by the Department of Engineering at the University of Cambridge in 2008.Segmentation and Recognition Using Structure from Motion Point Clouds".