Brief


Machine vision and the ‘Internet of Things’ (IoT) promise to revolutionise a wide range of everyday services. But the vision of smart buildings and smart cities will be hampered if it means replacing existing infrastructure and so requires significant new investment.

We joined forces with Analog Devices to set out to prove that machine vision and a low-cost processing platform could be harnessed to monitor car park occupancy and on-street parking – without expensive infrastructure. 

Approach


Our system combines a low-cost camera with a sophisticated algorithm to calculate which parking spaces are full or empty. It monitors occupancy over a wide area, without the disruption or expense of digging up roads and car parks to install individual sensors and communications for each parking bay.

The system is enabled by Analog Devices’ Blackfin® Low Power Imaging Platform – a low-cost, low-power embedded computer vision platform targeting a variety of real-time sensing applications. 

Benefit


Our algorithm ensures the technology functions effectively in different lighting conditions, with different vehicle sizes and in potentially misleading scenarios – such as when shoppers stand in a parking space. In the future, such a system could underpin the locating of vehicles, automated payment, parking space reservation and more.

Our solution shows how machine vision can ignite a range of IoT applications – providing low-cost, real-time sensing for smart buildings and smart cities.