One of the reasons that Miovision TrafficLink can count and classify cars, trucks, pedestrians and cyclists in real-time is that we do a lot of the needed computer processing right at the intersection.
In an age where many consumer and business applications are shifting to the cloud, this might seem counterintuitive. But, analysing data at the intersection avoids having to move bandwidth-heavy video somewhere else for processing, saving time and money. When TrafficLink is used to detect vehicles – for signal control, for example – analysing video in real-time avoids the latency involved with moving data hundreds or thousands of miles away for processing.
Of course, we’re not the only application for the growing Internet of Things (IoT). As more applications benefit from having local computing power in a small, robust package, chip-makers are creating new products to support so-called edge computing.
Chip-maker Nvidia, who is a supplier to Miovision, recently posted a blog: What is Edge Computing? that provides a nice overview of the concept and why it’s important for smart cities.