Using mobile phone to collect object data
To achieve spatial positioning of specific objects, we provide a mobile-based acquisition solution. If you need to perform spatial positioning and recognition of objects, you should use this feature. Before collecting data with your phone, please carefully read the following data collection instructions.
Acquisition preparation
Before starting data collection, you need to select objects suitable for visual positioning and prepare the corresponding software and hardware environment. Adequate preparation is key to obtaining high-quality map data. Before each collection, you should follow the Collection Checklist for preparation.
Suitable objects for positioning
It is recommended to use objects with rich textures for positioning. The specific positioning effect may vary depending on the object.
- Object types: statues, cars, billboards, chairs, toy figurines, and other objects with rich textures
- Object size: The target object should be larger than 10cm x 10cm x 10cm. Objects that are too small will result in reduced mapping quality and positioning accuracy
Selecting the appropriate phone model
Before starting collection, ensure your device meets the following requirements:
- iPhone users: iPhone 11 and above.
- Android users: Devices supporting ARCore (excluding devices disabled by EasyAR), view supported device details.
Note
Note that on ARCore-supported devices, the additional "Google Play Services for AR APP" must be installed to run ARCore functionality. This app may be pre-installed on some models, while on others, you need to search and install it from the app store.
Installing the collection tool
Download and install the data collection tool app Mega Toolbox:
iPhone users: Search for Mega Toolbox in the App Store, download, and install.
Android users: On the EasyAR download page, find the "Development Tools" section and download the Mega Toolbox Android installation package. After downloading, copy the installation package to your phone and install it directly, or use
adb install EasyARMegaToolbox_***.apkon your computer.
Acquisition process
The collection process is divided into three stages: launching Mega Toolbox, on-site multi-angle scanning collection, and data saving.
Launching Mega Toolbox
Open the installed
Mega Toolboxon your phone, click theEasyAR Mega Mapping Data Collection (Mobile Mapping)button to enter data collection mode.
Click the blue
+at the bottom to enter the collection page.
Function description
- Return to main interface
- Delete record
- Enter collection page
During the first collection, allow MegaToolbox to access your camera.

Performing collection
After entering the shooting interface, you need to follow a specific motion trajectory to scan the object from all angles to help build a complete visual map coverage.
Environment and initialization:
For movable objects, it is recommended to place them in a flat, open area with uniform lighting (such as a table); for immovable objects, clear temporary debris around the target object.
After entering the collection interface, first point the camera at a nearby area with rich textures and slowly pan to initialize. Once the status in the top-left panel changes to
Tracking, lift your phone towards the position to be collected and click the start collection button.
Multi-angle circular scanning:
With the object as the center, keep the phone camera facing the target object and perform
continuous circularcollection three to four times, specifically:- Collect one circle horizontally: Keep the camera height at the middle of the object and collect one circle horizontally.
- Collect one circle from a top-down view: Raise the phone and tilt the camera downward by 10~45 degrees to collect one circle in a top-down posture.
- Collect one circle from a bottom-up view: Lower the phone and tilt the camera upward by 10~45 degrees to collect one circle in a bottom-up posture. If the object is small, this step can be skipped.
- Top supplementary collection: If the object is small, collect once more from above the object.
Long-distance and detail supplementation:
- For larger objects, also collect one circle from a slightly farther distance, ensuring the object fills the entire camera frame as much as possible.
- If the object has many details or requires focused positioning in a specific area, collect more local perspectives of that part.
Tip
- Collection should be a single continuous session; do not interrupt during collection.
- During collection, the phone should be more than 0.5 meters away from the object. The collection distance depends on the object size but should ensure the collected data covers the entire object.
- The trajectory lines displayed in real-time during collection can help judge whether the collection complies with the requirements.
Ending and saving
After completing collection from all perspectives, you need to stop recording and save the data persistently to the device.
After collection is complete, click the red button in the middle at the bottom to stop collection and enter the save page.

On the save page, enter a name for the collected data (only English and numbers are supported), click the
Savebutton, and the collected data will be saved to the phone.
Precautions
- The entire collection process must maintain the
Trackingstate. Keep the collection smooth throughout, avoiding rapid turns or sudden stops. - The collection duration must not exceed
15 minutes. Exceeding15 minuteswill force the collection to end and enter the save page. - Do not switch to the background during collection, as this will cause collection to fail.
- Ensure the environment has sufficient, uniform, and stable lighting; avoid areas that are too dark or overexposed.
- Do not point the camera at plain surfaces like white walls or metal, or reflective/transparent areas like glass for extended periods.
- If collection fails, restart the collection.
Next steps
After collection is complete, refer to the Export Data document to export the data. Then, create a mobile object collection task in the Mega development center for mapping: Create Mobile Object Task.