Dragonfly Help > Working with Meshes > Registering Meshes

Registering Meshes

Dragonfly's iterative closest point (ICP) registration workflow can automatically register meshes by applying the rotation and translation that best aligns the selected mesh with a reference.

Right-click the required mesh in the Data Properties and Settings panel and then choose Mesh Registration in the pop-up menu to open the Mesh Registration panel, shown below.

Mesh Registration panel

Mesh Registration panel

Parameters available for Mesh Registration
  Description

Mobile mesh

Is the mesh that will be registered with the stationary reference mesh. You should note that you can register multiple meshes with the baseline, as well as change the baseline.

Stationary reference mesh

Is the baseline or reference mesh and will not be modified during the registration process.

Iterations

Allows you to select the maximum number of iterations that will be applied. Each iteration includes three main steps — finding the closest points, calculating the alignment, and updating the scene. You should note that if the registration criteria (tolerance) is met before the maximum number of iterations is performed, the process will stop.

Tolerance

Lets you set the required accuracy of the registration at each iteration of the ICP.

Rejection scale

Lets you set the criteria for rejecting outliers, which can increase the robustness of registrations. This value corresponds to the standard deviation coefficient and points within the Rejection scale will be ignored during registration.

Number of levels

Lets you set the number of pyramid levels. Deep pyramid levels may increase speed, but can decrease accuracy, while coarse pyramids might have computational overhead on top of inaccurate registration. This parameter should be chosen to optimize a balance. Typical values range from 4 to 10.

Refer to https://docs.opencv.org/3.3.1/dc/d9b/classcv_1_1ppf__match__3d_1_1ICP.html for information about the implementation of ICP in Dragonfly. Refer to https://en.wikipedia.org/wiki/Iterative_closest_point for general information about the iterative closest point algorithm.

 

Dragonfly Help Live Version