Mega diagnostic data collection
This guide aims to provide standardized guidance for collecting visual localization anomaly data. Through standardized data collection and feedback, it assists developers in accurately describing issues and helps the EasyAR technical team efficiently locate, reproduce, and resolve technical faults related to visual localization.
Content overview
During the integration of EasyAR Mega localization services, accurate data collection is the core prerequisite for problem diagnosis. This guide covers the following key aspects:
- Issue identification: Classification and preliminary assessment of common abnormal phenomena.
- Screenshot/screen recording specifications: Visual evidence collection methods for different device platforms.
- Runtime data collection: Extraction and export of runtime environment data across platforms.
- Information feedback standards: Ensuring the "minimum information set" required by the technical team to reproduce issues.
Classification and preliminary assessment of abnormal phenomena
Before reporting issues, please categorize abnormal phenomena according to the following classifications:
Common abnormal manifestations
- Complete localization failure: Unable to recognize scenes, status persistently displays as "NotFound".
- Localization accuracy deviation: Successful localization but insufficient accuracy, with noticeable spatial offset.
- Unstable localization: Visible jitter, drift, or frequent tracking loss.
- Regional anomalies: Poor localization performance in specific areas or points.
- Device compatibility differences: Suboptimal performance on certain specific device models.
Possible cause speculation
- Scene-related: Content placement offset, poor mapping quality, sparse environmental lighting or feature points (e.g., large white walls), interference from similar points.
- Technical environment: Network jitter, unpaid service fees, QPS limits, SDK version compatibility, API call logic errors.
- Hardware permissions: Device sensors (IMU) not supporting motion tracking, camera or storage permissions not enabled.
Standardized feedback process
To improve troubleshooting efficiency, we strongly recommend submitting a complete feedback package containing the following three components:
Basic information checklist (minimum information set)
- Resource identifiers: Account information, Mega AppId, Block Id.
- Version information: EasyAR Sense SDK version or mini-program plugin version.
- Problem details: Detailed description of abnormal phenomena and their trigger conditions.
- Environmental parameters: Device model, OS version, WeChat version, current network environment (4G/5G/WiFi).
Visual evidence (video and screenshots)
- Reproduction process recording: Complete recording of the localization process from startup to anomaly occurrence.
- Keyframe screenshots: Capture instant snapshots when anomalies occur.
- Multi-angle recording: Provide visual references under different positions and lighting conditions.
Runtime data (core troubleshooting basis)
- EIF data: Runtime records based on EasyAR Sense, supporting playback analysis on PC.
- Dump data: WeChat mini-program specific runtime state snapshots.
- Application logs: App runtime logs (Logcat/Console) and error stacks.
Tip
Further reading: How to clearly communicate problem details
Runtime environment and collection method classification
Data collection methods vary depending on the runtime device and platform. Please select the corresponding guidance based on your actual environment.
Device classification description
- Mobile devices: Smartphones (iOS/Android).
- Head-mounted devices:
- OST (Optical See-Through): Optical see-through (e.g., XREAL, Rokid).
- VST (Video See-Through): Video see-through (e.g., Apple Vision Pro, PICO)
Screenshot and screen recording guidance
| Platform type | Reference guide |
|---|---|
| Mobile / WeChat mini-program | Mobile screenshot/screen recording and export |
| OST head-mounted | OST screen recording and export |
| VST head-mounted | VST screen recording and export |
Runtime data collection (EIF/Dump)
| Scenario type | Collection solution |
|---|---|
| EasyAR Sense native application | Mobile app EIF recording / Head-mounted EIF recording |
| WeChat mini-program | Mini-program dump data recording and export |
| Quick collection tools | Mega Toolbox EIF recording / Toolbox dump recording |
Automatic collection feature (coming soon)
We will soon upgrade to support automatic collection and feedback features, providing a more convenient data submission method:
- Background recording: Start background collection of required datasets on demand
- Automatic upload: Automatically package and upload to the EasyAR technical team
- On-demand capture: Self-service real-time anomaly state capture
- Progress tracking: Categorization and continuous tracking
Operation guidance navigation summary
Jump directly to detailed operation guides based on your device platform requirements:
| Type | Screen recording/screenshot collection guide | Runtime data collection guide |
|---|---|---|
| Mobile application | Mobile screenshot/screen recording and export | Mobile app EIF recording, Mega Toolbox EIF recording, EIF data export instructions |
| OST head-mounted | OST screen recording and export | Head-mounted device EIF recording and export |
| VST head-mounted | VST screen recording and export | Head-mounted device EIF recording and export |
| WeChat mini-program | Mobile screenshot/screen recording and export | WeChat mini-program dump recording and export, WeChat mini-program Toolbox dump recording |
Best practice recommendations
- Pre-research process: Before project launch, team members should familiarize themselves with the data collection process.
- Self-check first: Before feedback, confirm whether technical environment issues, content placement problems, or device permission issues have been excluded.
- Data completeness: Collect complete data in a standardized manner: feedback minimum information set, visual anomaly videos and screenshots, runtime data.
- Timely communication: After submitting complete data, maintain communication with the EasyAR technical team.
Technical support channels
- Help documentation: Consult Mega documentation or Troubleshooting.
- Technical support: Contact us immediately or email support@sightp.com.
Next topic: Mobile screenshot/screen recording and export methods