Our proposal would be to keep the implementation as is and to update the description. Further, we propose to allow a negative value as a lower bound on the "lower_threshold" to allow no filtering on the lower bound. With the current implementation we saw in simulation that laser measurements with zero intensity were filtered out as well, which is undesired.
As an alternative, we can keep the dynamic configuration as is, but make the implementation consistent by replacing the "<=" statements in the link above by "<".
The "lower_threshold"-param of the intensity filter describes that "Intensity values lower than this value will be filtered", while the implementation filters for values "lower equal", see https://github.com/ros-perception/laser_filters/blob/noetic-devel/src/intensity_filter.cpp#L78
Our proposal would be to keep the implementation as is and to update the description. Further, we propose to allow a negative value as a lower bound on the "lower_threshold" to allow no filtering on the lower bound. With the current implementation we saw in simulation that laser measurements with zero intensity were filtered out as well, which is undesired.
As an alternative, we can keep the dynamic configuration as is, but make the implementation consistent by replacing the "<=" statements in the link above by "<".