publish poses at the end of scan

[koide3]

No description provided.

[Copilot]

Pull request overview

Adds dedicated ROS2 topics to publish the estimated IMU pose/odometry at the end of a LiDAR scan, in addition to the existing per-frame pose/odom publications, to improve RViz visualization and downstream consumers that need scan-end timing.

Changes:

• Add new publishers for scan-end odometry/pose, with both uncorrected and corrected variants.
• Compute scan-end relative motion from imu_rate_trajectory and publish /odom_scanend* and /pose_scanend* messages using the scan-end timestamp.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 4 comments.

File	Description
src/glim_ros/rviz_viewer.cpp	Adds scan-end time/transform computation and publishes new scan-end odom/pose topics.
include/glim_ros/rviz_viewer.hpp	Declares the new scan-end publishers (regular + corrected).

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The scan-end warning condition looks backwards/noisy: std::abs(imu_end_time - new_frame->stamp) < 1e-3 will be true whenever imu_rate_trajectory is empty (or has a single sample), causing a WARN every frame while still publishing a scan-end pose identical to the frame pose. Consider warning only when imu_rate_trajectory is missing (or when the time delta is unexpectedly large), and/or downgrade to DEBUG/throttle so normal operation doesn’t spam logs.

Suggested change

	if (std::abs(imu_end_time - new_frame->stamp) < 1e-3) {
	logger->warn("Scan end time is too close to the frame time (imu_end_time={}, frame_time={})", imu_end_time, new_frame->stamp);
	if (std::abs(imu_end_time - new_frame->stamp) > 1e-3) {
	logger->warn("Scan end time differs significantly from the frame time (imu_end_time={}, frame_time={})", imu_end_time, new_frame->stamp);

[Copilot]

std::abs is used here for doubles, but this file doesn’t include <cmath>. Relying on transitive includes can break builds depending on compiler/standard library; please include <cmath> explicitly (or switch to std::fabs).

[Copilot]

The debug log after publishing scan-end odometry logs new_frame->stamp and uses the same message as the regular odom publisher. This is misleading because the published message is timestamped with imu_end_stamp; consider logging imu_end_time (and differentiating it as scan-end odom) so users can diagnose timing correctly.

Suggested change

	logger->debug("published odom (stamp={})", new_frame->stamp);
	logger->debug("published scan-end odom (imu_end_time={}, stamp={})", imu_end_time, imu_end_stamp);