Data Creation and Movement
Hivemapper-supported dashcams target capturing high-quality imagery (2k-4k) at 10 frames per second. Additionally, dashcams capture high-precision GNSS and IMU data. The Open Dashcam API then transfers the images and associated sensor data to the mobile device via the Hivemapper contributor app.
The Hivemapper app plays an important role in collecting the images to be sent to the Hivemapper Network. On average, images are collected every 8 meters. The mobile device packages these images in bundles, which are used to send information through the system in a more efficient manner. Typically, these bundles include anywhere from 10-100 images at a time. Images that don’t meet the network’s usability standards are filtered out (e.g., poor lighting conditions, unverifiable location). When the app has access to upload data, it does so in bundles of images as well.
Once data is uploaded to the Hivemapper Network, it enters a queue to be processed. Data from contributors with bad reputation scores is deprioritized in the processing queue.
Processing imagery involves verifying the image is not corrupt, the image and its location are authentic, and that the image is high quality. Additional quality checks related to mount position, blurriness, glare, obstructions, etc., are also completed.
After the data is verified by the system, images pass through a manual quality assurance process performed by contributors.