feat(multi_object_tracker): add object class filtering in tracking process

[technolojin]

Description

The current object tracker uses the incoming classification directly.
When the detection has ambiguous classification, the classification result may changes in a short time.
Frequent classification change may have negative impact on the tracked object consumers (planner).

In this PR, a simple classification filter is introduced.

// 0. Normalize the input classification
// 1-1. Update the matched classification probability with a gain (ratio of 0.05)
// 1-2. If the label is not found, add it to the classification list
// 2. Remove the class with probability < remove_threshold (0.001)
// 3. Normalize tracking classification

Related links

TIER IV INTERNAL

Tests performed

xx1-class-filter-implemented-bicycle-pedestrian.mp4

Blue: Detected Object (input)
Yellow: Tracked Object (output)

The result shows the highest probable label is change only when the detection label is changed for a while.

Notes for reviewers

This PR do not improve performance of detection classification.

Interface changes

Not Applicable

Effects on system behavior

Not Applicable

Pre-review checklist for the PR author

The PR author must check the checkboxes below when creating the PR.

• I've confirmed the contribution guidelines.
• The PR follows the pull request guidelines.

In-review checklist for the PR reviewers

The PR reviewers must check the checkboxes below before approval.

• The PR follows the pull request guidelines.
• The PR has been properly tested.
• The PR has been reviewed by the code owners.

Post-review checklist for the PR author

The PR author must check the checkboxes below before merging.

• There are no open discussions or they are tracked via tickets.
• The PR is ready for merge.

After all checkboxes are checked, anyone who has write access can merge the PR.

[YoshiRi]

Cool! Could you tell me your opinion about my questions?

[YoshiRi]

LGTM. It worked within my environment.