Add intensity for poincloud processing in perception and an intensity_based_unknown_validator

[badai-nguyen]

Checklist

• [X] I've read the contribution guidelines.
• [X] I've searched other issues and no duplicate issues were found.
• [X] I've agreed with the maintainers that I can plan this task.

Description

• Durring and after raining weather, autoware faces a issue of false detection of Unkown caused by pointcloud from raidrop or water splash from ego and other fast moving vehicle.

Purpose

• Reduce or ignore such noise pointcloud uknown detection.

Possible approaches

In order to reduce or ignore the unknown of such noise pointcloud, there are two possible appoaches

• Apply some kind of noise pointcloud denoise filter during pointcloud processing stage
• Implement a kind of unknown object validator in object recognition stage (after Euclidean Cluster).

This task is to plan the implementation of an intensity based unknown validator for the later approach.

Definition of done

• [ ] Implement the intensity_based_validator node
• [ ] Update and add intensity field for the neccesary node for above validator

Currently I think the nodes which doesn't have intensity yet and needed to be added are

• [ ] Scan Ground Filter
• [x] occupancy_grid_map_outlier_filter
• [x] compare_map_segmentation
• [x] voxel_grid_downsample_filter
• [ ] euclidean_cluster
• [ ] ~Shape_estimation~
• [ ] ~Detection_by_tracker~
• [x] roi_cluster_fusion_node