Motionloft Blog

Computer Vision at the Edge by Motionloft

Posted by Panos Kourdis on Dec 3, 2018 11:21:04 AM
Find me on:

Motionloft's weatherized ViMo sensor provides in real-time accurate pedestrian and vehicle tracking information inside or outside, visualized via our in-house built web-analytics platform and also accessible via a secure private API. It connects to the Cloud wirelessly by LTE or WLAN without requiring any additional infrastructure. Only, a common 110 volt electrical outlet and a good spot to mount the sensor for capturing the region of interest is sufficient.


ViMo is digitizing its observable environment via a dual-camera setup enabling stereoscopic vision, which is processed by an on-board NVIDIA GPU for AI inference at the edge. Our in-house developed C++ Computer Vision application is running optimized for deployment and inference at the edge CNN object detection models combined with advanced geometric path tracking algorithms to accurately map the object's motion in 3D space.

Tags: pedestrian Sensor, Pedestrian Counts, Computer Vision, vehicle counts, pedestrian analytics, vehicle analytics


We Turn Video Into Meaningful Data

Motionloft takes the physical movement of people and vehicles and turns it into digitized data. Our dashboard presents this unique data to the user in real-time enabling them to make confident decisions, provide precision for algorithms, and reveal new opportunities that only data can surface.


Subscribe to Email Updates

Recent Posts