Advanced Computer Vision
We employ state-of-the-art algorithms for parking space classification and have large, consistent datasets – allowing us to overcome the challenges traditionally associated with vision-based systems. Our technical advantages provided an excellent opportunity to implement advanced deep learning techniques for parking lot recognition.
Our algorithms surpass the latest research publication demonstrating superhuman accuracy of convolutional neural network (CNN) algorithms in cross-parking occupancy recognition, in a variety of lighting and weather conditions. Real-time deep learning occupancy recognition is estimated to be a very efficient and inexpensive solution for increasing the parking experience of drivers, as well as reducing the costs of facility management.
Moreover, connecting parking occupancy and license plate recognition (LPR) of each vehicle with our parking facility management dataset of registered cars for a network of cameras can provide real-time automated permit checks of parked vehicles at a fraction of the cost of traditional parking monitoring solutions. Furthermore, by time stamping the parked vehicles, the system can calculate the occupied time of every parked vehicle, and recognize the time limit of a specific spot, if necessary.
Standalone Edge Sensors
Visionful Edge unit monitors 360-degree view powered by a dedicated GPU and our AI software. These standalone and portable units can work with solar panels and has its own IoT cellular connection, removing the need for power and network infrastructure. The Edge computation eliminates the need for transferring large files to the cloud. Our product is designed to meet the IP68 standard, providing excellent water and dust protection.
The camera’s flexible variable mounting allows being installed in different ways to accommodate for different directions. Flexibilities in the use of power and embedded cellular connectivity made installation simple and cost-efficient.
Each camera is connected to our server through an internal IOT Cellular module. It resolves the complications of building a network infrastructure at the site.
Weather is subject to severe and constantly change in conditions, our cameras have been tested to perform under the most challenging weather under I68 dust and water standards.
To reduce costs and promote energy sustainability, our cameras have the capabilities to use solar technologies. They can be connected to constantly powered or timed lighting poles as well.
Powerful Nvidia Graphic Processing Unit (GPU) inside each sensor equipped our hardware with the same technology used in self-driving cars resulting in outstanding accuracy.
Embedding the Computer vision software on our sensors makes it possible to perform all the calculations on the edge resulting in a seamless process and a massive saving on data transfer and server costs.