Home

Welcome to the Depth-Preception wiki!

The first steps of the project must consist of research on current state of the art depth preception.

The second step will be to attain sufficient data that contrains LIDAR data and RGB images.

Third; must find a way to associate each image to a set of LIDAR data points. This will use the sementation data that should be provided. [Must be an automated process]

Fourth; use state of the art mono-depth preception using multiple frames. This will make the inital approximations of the depth of the image. Using some linear regression model we update these depth predictions. Update these parameters and continue to update until convergence.