The two main issues that we may face when attempting to process point clouds are excessive noise and excessive density. The former causes our algorithms to misinterpret the data and produce incorrect or inaccurate results, whereas the latter makes our algorithms take a long time to complete their operation. In this section, we will provide insight into how to reduce the amount of noise or outliers of our point clouds and how to reduce the point density without losing valuable information.
The first part is to create a node that will take care of filtering outliers from the point clouds produced in the pcl_output topic and sending them back through the pcl_filtered topic. The example can be found in the source directory ...
