VL Task

Available in Classic and VPC

VL Task describes how to convert the space image data you collected into location data.

The ARC eye location data conversion consists of the following 4 stages:

1. Convert to SLAM data: (Lowest layer) Convert the spatial image data into SLAM data.
2. Merge SLAM data: (Middle or top layer) Merge SLAM data you converted in the lowest layer.
3. Localize: (Top layer) Localize the SLAM data you used to merge making it location data.
4. Merge into VL data: (Top layer) Merge the SLAM data into API serviceable VL data.

VL Task screen

The VL Task interface includes the following components:

Lowest stratum screen

The functions at the lowest stratum are as follows:

Middle stratum screen

The functions at the middle stratum are as follows:

Highest stratum screen

The functions at the highest stratum are as follows:

View VL task list

You can view the task list and further information. To view the information:

1. Navigate to > Services > Digital Twin > ARC eye in the NAVER Cloud Platform console.
2. Navigate to Visual Localization > VL Task .
3. Click the name of VL task (segment) in the location tree that you want to view the information.
4. View the summary in the VL task list or click a task to view the detailed information.
   
   • Task type: It displays the applicable task type among SLAM data conversion, SLAM data merge, localization, and VL data conversion.
   • Task name: The name of the task you created
   • Creation date and time (UTC+09:00): Date and time when the task was created
   • Completion date and time (UTC+09:00): Date and time when the task was completed
   • Server duration: How many hours the server was used for the task
   • Progress: Displays the progress rate of the task (%)
   • Verification: Verification and approval of the task completion. The Merge VL data task do not require the validation stage.
   • Result: Task progress status
     • In progress: Task in progress
     • Need validation: Validation and approval regarding the task completion must be done at Potree Viewer. (Click in the Validation row.)
     • Succeeded: Task succeeded
     • Failed: Task failed
   • Segment name: The segment name of a location you created at the Location menu
   • Task completion list: Displays the task name applicable for data merge or conversion

Convert SLAM data

To use ARC eye, you must convert the spatial image data you collected into SLAM data. SLAM data conversion is only available in the lowest layer. To convert spatial image data into SLAM data:

1. Navigate to > Services > Digital Twin > ARC eye in the NAVER Cloud Platform console.
2. Navigate to Visual Localization > VL Task .
3. Select a project you want from the Project name dropdown list.
4. Click the lowest stratum in the Location Tree and undergo the following stages in order.
   • 1. Upload spatial image data
   • 2. Convert SLAM data
   • 3. Validate SLAM conversion data

1. Upload spatial image data

Upload the spatial image data the user collected to the lowest layer to convert it to SLAM data. To upload them:

1. Click [Upload spatial image data] .
2. When the Upload image data window appears, click [Upload] .
3. When the Upload pop-up window appears, drag the files into the Drag and drop the file for upload. area, or click it to select files to upload.
   • You can delete unnecessary files by clicking in the upload target list.
4. Once you are ready to upload, click [Upload] .
   • To cancel during the data upload, click [Stop transfer] .
5. View the Upload image data screen if the task is added.
   • To delete a task, select the task, click [Delete selection] and [Delete] in the Delete selection pop-up window.
   Caution

   Upon deleting a task, all the stored data of the task is deleted. Please decide carefully.

2. Convert SLAM data

2.1. Run SLAM data conversion

Convert the uploaded spatial image data to SLAM data. To convert data:

1. Click to go to the VL Task page.
2. Click [Convert SLAM data] .
3. When the SLAM data conversion pop-up window appears, enter the following information and click [Run] .
   • Task name: Enter the name of the task. You can edit the task name instead of using the default task name.
   • Target task list: The names of the space image data the user uploaded
   • Space image preset: Select the preset you want in the dropdown list
     • If the mapped target range is over 3500 ㎡ (usually outdoors), select a preset with large in its name. If it is smaller (usually indoors), select a preset with small in its name (recommended). If you select a preset fitting to your situation, more appropriate algorithm is used for the process and the Positioning API’s accuracy increases.
   • The task is added to the task list and the progress rate is shown.
   • If you have set an alarm, the notification recipient receives the notification upon completion of the task. For more information on setting notification recipients, see Notification Setting.

2.2. SLAM data conversion result

After the task is executed, you can check its status via the progress indicator, and the following results are displayed for the processing.

3. Validate SLAM conversion data

Once the data conversion is done, you need to verify if the collected data was converted correctly. You can check whether the space has been implemented correctly in Potree Viewer. To verify converted data:

1. Click .
2. When the Potree Viewer popup window appears, apply the set value on the left setting window to verify.
3. Once the data validation is done, click [Approve] .
   • The result of the task in the task list changes from Need validation to Succeeded.
   • Click [Cancel] to change that task's result to Failure.
   • If you need a reexamination due to data abnormality and such, click [Review again] to exit the Potree Viewer and proceed with the necessary task.

4. Response measures

• If there are problems with the inspection data, follow these response measures in order.
• Response measure: Select and run a false loop preset (e.g., Small false loop)
  • If it fails, select the insufficient loop preset and run it.
  • If it fails, select the registration error preset and run it.
  • If it fails, select the odometry error preset and run it.

Merge SLAM data

After converting the spatial image data in the lowest layer into SLAM data, you must merge the converted SLAM data in all layers ascending from the lowest layer. To merge SLAM data:

1. Navigate to > Services > Digital Twin > ARC eye in the NAVER Cloud Platform console.
2. Navigate to Visual Localization > VL Task .
3. Select a project you want from the Project name dropdown list.
4. Click the target segment in the Location Tree and undergo the following stages in the order.
   • 1. Merge SLAM data
   • 2. Validate merged SLAM data

1. Merge SLAM data

1.1. Run SLAM data merge

To merge data:

1. Click [Merge SLAM data] .
2. When the SLAM data merge pop-up window appears, enter the following information and click [Run] .
   • Segment name: Displays the selected segment
   • Task name: Enter the name of the task. You can edit the task name instead of using the default task name.
   • Target task list: Select SLAM data.
     • Upon merging multiple data together, position them in the order considering the connection between data.
     • Map merging proceeds sequentially from the parent location of the location where SLAM data conversion was performed.
     • For example, if there are four total layers, after executing SLAM data conversion for Layer 4, it sequentially proceeds with SLAM data merging in Layer 3 → SLAM data merging in Layer 2 → SLAM data merging in Layer 1.
     • Ensure all data required for SLAM data merging is uploaded and SLAM data conversion is complete before proceeding. For example, if five data sets are needed, do not merge only three first.
   • Space image preset: Select the preset you want in the dropdown list
     • If the mapped target range is over 3500 ㎡, select a preset with large in its name. If it is smaller, select a preset with small in its name (recommended). If you select a preset fitting to your situation, more appropriate algorithm is used for the process and the Positioning API’s accuracy increases.
   • The task is added to the task list and the progress rate is shown.
   • If you have set an alarm, the notification recipient receives the notification upon completion of the task. For more information on setting notification recipients, see Notification Setting.

1.2. SLAM data merge result

After the SLAM data merge task is executed, you can check its status via the progress indicator, and the following results are displayed for the processing.

2. Validate merged SLAM data

Once the SLAM data merge is done, you need to validate if the collected data was merged correctly. You can check whether the space has been implemented correctly in Potree Viewer. To validate:

1. Click .
2. When the Potree Viewer popup window appears, apply the set value on the left setting window to verify.
3. Once the data validation is done, click [Approve] .
   • The result of the task in the task list changes from Need validation to Succeeded.
   • Click [Cancel] to change that task's result to Failure.
   • If you need a reexamination due to data abnormality and such, click [Review again] to exit the Potree Viewer and proceed with the necessary task.

4. Response measures by SLAM data merge validation data

• If there are problems with the inspection data, follow these response measures in order.
• Response measure: Select and run a false loop preset (e.g., Small false loop)
  • If it fails, select the insufficient loop preset and run it.

Localization

After merging all SLAM data, run the localization at the top stratum. Localization is a task to map the accurate location information by unifying the location values of the merged SLAM data. This is to enable multiple datasets to share a common coordinate system. To localize the merged SLAM data:

1. Navigate to > Services > Digital Twin > ARC eye in the NAVER Cloud Platform console.
2. Navigate to Visual Localization > VL Task .
3. Select a project you want from the Project name dropdown list.
4. Click the top target segment in the location tree and proceed with the following stages in order:
   • 1. Localize data
   • 2. Validate localized data

1. Localize data

1.1. Run data localization

If you have a successful SLAM data merge operation at the top layer, you can proceed to localize. To localize data:

1. Click [Localize] .
2. When the Localize pop-up window appears, enter the following information and click [Run] .
   • Segment name: Displays the selected segment
   • Task name: Enter the name of the task. You can edit the task name instead of using the default task name
   • Task target list: Select the merged SLAM data.
   • Space image preset: Select the preset you want in the dropdown list
     • If the mapped target range is over 3500 ㎡, select a preset with large in its name. If it is smaller, select a preset with small in its name (recommended). If you select a preset fitting to your situation, more appropriate algorithm is used for the process and the Positioning API’s accuracy increases.
   • The task is added to the task list and the progress rate is shown.
   • If you have set an alarm, the notification recipient receives the notification upon completion of the task. For more information on setting notification recipients, see Notification Setting.

1.2. Data localization result

2. Validate localized data

Once the localization is done, you need to validate if the data was localized correctly. You can check whether the space has been implemented correctly in Potree Viewer. To validate:

1. Click .
2. When the Potree Viewer popup window appears, apply the set value on the left setting window to verify.
3. Once the data validation is done, click [Approve] .
   • The result of the task in the task list changes from Need validation to Succeeded.
   • Click [Cancel] to change that task's result to Failure.
   • If you need a reexamination due to data abnormality and such, click [Review again] to exit the Potree Viewer and proceed with the necessary task.

4. Response measures by Localized data validation result

• If there are problems with the inspection data, follow these response measures in order.
• Response measure: Select the registration error preset and run it. Example: Small registration error
  • If it fails → Select the false loop preset and run it.
  • If it fails → Select the insufficient loop preset and run it.

Validate data

Proceed through each stage, from SLAM data conversion and SLAM data merging to localization, while reviewing the data status via the Potree Viewer. In addition to the review methods for each stage, also refer to the types of common issues that may arise.

• If it overlaps or duplicates incorrectly

Failure	Succeeded
The path is distorted, and the map overlaps in an incorrect form.	-
It overlaps in an incorrect form.	-

• If blurred or hazy
  

VL

1. Run VL

You can convert SLAM data into VL data suitable for API services. This task is only available in the top stratum. To convert SLAM data into VL data:

1. Navigate to > Services > Digital Twin > ARC eye in the NAVER Cloud Platform console.
2. Navigate to Visual Localization > VL Task .
3. Select a project you want from the Project name dropdown list.
4. Click the target segment of the highest stratum in the location tree.
5. Click [VL] .
6. When the VL pop-up window appears, enter the following information and click [Run] .
   • Segment name: Displays the selected segment
   • Task name: Enter the name of the task. You can edit the task name instead of using the default task name.
   • Target task list: Select the converted SLAM data.
   • The task is added to the task list and the progress rate is shown.
   • If you have set an alarm, the notification recipient receives the notification upon completion of the task. For more information on setting notification recipients, see Notification Setting.

2. VL result

3. Delete VL

You can delete the registered VL tasks. To delete it:

1. Navigate to > Services > Digital Twin > ARC eye in the NAVER Cloud Platform console.
2. Navigate to Visual Localization > VL Task .
3. Select a project you want from the Project name dropdown list.
4. Click the segment you wish to delete tasks from in the location tree.
5. Click and select the task you want to delete in the segment’s task list and click [Delete selection] .
6. Once the Delete selection pop-up window appears, click [Delete] .

VL data merge

1. Run VL data merge

You can convert SLAM data into VL data suitable for API services. This task is only available in the top stratum. Once VL data conversion is done, API is created automatically. You can start your service by deploying this API. To merge VL data:

1. Navigate to > Services > Digital Twin > ARC eye in the NAVER Cloud Platform console.
2. Navigate to Visual Localization > VL Task .
3. Select a project you want from the Project name dropdown list.
4. Click the target segment of the highest stratum in the location tree.
5. Click [Merge VL data] .
6. When the Merge VL data pop-up window appears, enter the following information and click [Run] .
   • Segment name: Displays the selected segment
   • Task name: Enter the name of the task. You can edit the task name instead of using the default task name.
   • Target task list: Select the converted VL data.
   • The task is added to the task list and the progress rate is shown.
   • If you have set an alarm, the notification recipient receives the notification upon completion of the task. For more information on setting notification recipients, see Notification Setting.

2. VL data merge result