Visualize SLAM Poses using SimpleViz - accumulates track, point clouds and map views

Overview

Beginning in ouster-sdk 0.13.0, it’s easier than ever to use “map” or “scan” accumulation to visualize lidar scans that contain pose information. The default visualizer, SimpleViz, present in both ouster-cli and Ouster SDK’s Python API support these accumulation modes out of the box.

Furthermore, when poses are not present in the LidarScan accumulation may still be useful to view the accumulated N scans from the live sensor/recording and reveal the accuracy/repeatability of the data.

Slam Viz command

The viz command enables visualizing the accumulated point cloud generation during the SLAM process. By default, the viz operates in looping mode, meaning the visualization will continuously replay the source file.

ouster-cli source <SENSOR_HOSTNAME> / <FILENAME> slam viz

When combining the viz and save commands, the saving process will automatically terminate after the first iteration, and then the SLAM process restarts for each subsequent lidar scan iteration. To end the SLAM and visualization processes after the save operation completes, you can use ctrl + c. Alternatively, you can add -e exit to the viz command to terminate the process after a complete iteration.

ouster-cli source <SENSOR_HOSTNAME> / <FILENAME> slam viz -e exit save sample.osf

Accumulation: The viz command supports several options for creating visually-pleasing maps by accumulating data from lidar scans that contain pose information. The following sections describe the options and provide usage examples.

Available view modes

There are three view modes of accumulation implemented in the default visualizer that may be enabled/disabled depending on its parameters and the data that is passed through it:

• poses mode - all scan poses in a trajectory/path view (if poses data is present in scans). key 8 can be used to toggle the path in viz.

• map accumulation mode - overall map view with select ratio of random points from every scan (available for scans with or without pose). key 7 can be used to toggle the map view.

• scan accumulation mode - accumulated N scans (key frames) that is picked according to parameters (available for scans with or without poses). key 6 can be used to toggle the scan accumulation view.

Ouster CLI viz accumulation options

• scan accumulation options

  • --accum-num INTEGER Accumulate up to this number of past scans for visualization. Use <= 0 for unlimited. Defaults to 100 if --accum-every or --accum-every-m is set.

  • --accum-every INTEGER Add a new scan to the accumulator for every specified number of scans as an argument in this option.

  • --accum-every-m FLOAT Add a new scan to the accumulator after specified number of meters of travel.

• map accumulation options

  • --map If set, add random points from every scan into an overall map for visualization. Enabled if either --map-ratio or --map-size are set.

  • --map-ratio FLOAT Fraction of random points in every scan to add to overall map (0, 1]. [default: 0.01]

  • --map-size INTEGER Maximum number of points in overall map before discarding. [default: 1500000]

Key bindings

The following key shortcuts apply to accumulation options while running Ouster CLI’s viz command.

Key	Accumulation Mode	What it does
6	Scan	Toggle scans accumulation view mode
7	Map	Toggle overall map view mode
8	Scan & Map	Toggle poses/trajectory view mode
k / K	Scan & Map	Cycle point cloud coloring mode of accumulated clouds or map
g / G	Scan & Map	Cycle point cloud color palette of accumulated clouds or map
j / J	Scan & Map	Increase/decrease point size of accumulated clouds or map

Dense accumulated clouds view (with every point of a scan)

To obtain the densest view use the --accum-num N --accum-every 1 parameters where N is the number of clouds to accumulate (N up to 100 is generally small enough to avoid slowing down the viz interface.)

The following example computes poses for each scan using the slam command and creates a dense map using the viz --accum-num 20 to accumulate the points from 20 scans. Finally, the save command writes the scans with their computed trajectories to an OSF file. (Note - accumulation is a visualization feature only. The accumulated data is not saved to the file.):

ouster-cli source <SENSOR_HOSTNAME> / <FILENAME> slam viz --accum-num 20 save sample.osf

and the dense accumulated clouds result:

Dense view of 20 accumulated scans during the slam viz run

Overall map view (with poses)

One of the main tasks we frequently need is a preview of the overall map. We can test this by using the SLAM-generated OSF file, which was created with the above command and contains the SLAM trajectory in LidarScan.pose. If you are using a SLAM-generated OSF, you can directly use viz with scan accumulator feature without appending the slam option.

ouster-cli source ouster_sensor_recording.osf viz --accum-num 20 \
--accum-every 0 --accum-every-m 10.5 --map -r 3 -e stop

Here is a preview example of the overall map generated from the accumulated scan results. By utilizing the ‘-e stop’ option, the visualizer stops once the replay process finishes, displaying the preview of the lidar trajectory:

Data fully replayed with map and accum enabled (last current scan is displayed here in gray palette)

Data fully replayed with view only last 20 scans accumulated every 10.5 meters

Data fully replayed with view of only trajectory (yellow knobs is 20 accumulated key frames positions)

Programmatic use

To use any of these accumulation modes, provide their configuration directly to SimpleViz via keyword arguments. The following snippet will play back scans from the source scans_w_poses and the sensor configuration provided by meta:

import sys
from ouster.sdk import open_source
from ouster.sdk.viz import SimpleViz
from ouster.sdk.mapping import SlamConfig, SlamEngine

source_uri = sys.argv[1]
source = open_source(source_uri)
config = SlamConfig()
slam = SlamEngine(source.sensor_info, config)

def scans_w_poses():
    for scan in source:
        yield slam.update(scan)


viz = SimpleViz(
    source.metadata,
    accum_max_num=100,
    accum_min_dist_num=0,
    accum_min_dist_meters=4,
    rate=1,
    on_eof='stop'
)

viz.run(scans_w_poses())

Alternatively, LidarScanViz (which is a lower-level visualizer that implements SimpleViz) can display a static map from scans that have poses computed in a preprocessing step:

import sys
from tqdm import tqdm  # for progress bar
from ouster.sdk import open_source
from ouster.sdk.viz import LidarScanViz
from ouster.sdk.viz.accumulators_config import LidarScanVizAccumulatorsConfig
from ouster.sdk.mapping import SlamConfig, SlamEngine

source_uri = sys.argv[1]
source = open_source(source_uri)
config = SlamConfig()
slam = SlamEngine(source.sensor_info, config)

num_scans_to_map = 200
scans_w_poses = [
    slam.update(scan) for _, scan in
    zip(tqdm(range(num_scans_to_map), desc="Computing map"), source)
]

viz = LidarScanViz(
    source.metadata,
    accumulators_config = LidarScanVizAccumulatorsConfig(
        accum_max_num=100,
        accum_min_dist_num=0,
        accum_min_dist_meters=4
    )
)

for scan in scans_w_poses:
    viz.update(scan)

viz.draw(update=True)
viz.run()