Doctoral Meeting: 'LitS: A novel Neighborhood Descriptor for Point Clouds'

In this talk I will discuss LitS, the object of study in the research paper we are currently working on. LitS are a point cloud feature conveying geometric information about each point's surroundings. Mostly angular, this information can be used to identify corner points, boundaries, and to discriminate otherwise singular points within the point cloud. Our paper delves into the mathematical underpinnings of LitS, focusing on its formulation for both 2D and 3D point clouds and elucidating its key properties. Furthermore, we demonstrate how LitS can be effectively leveraged for boundary detection tasks, and we provide illustrative examples that showcase LitS's ability to discriminate points across various parameter settings.

Supervisors

• Francisco Fernández Rivera
• Óscar García Lorenzo

Moderator of this Doctoral Meeting: Manuel Mucientes Molina