Video Analytics From MultiSight: All Value, No Fuss

By Steve Mitchell, Director of Engineering, MultiSight

In designing features like heatmaps and people counting, we exploit opportunities of MultiSight’s overall architecture in ways many other video systems cannot, so that we can offer these features at lower cost and with greater flexibility.

MultiSight is a distributed system, so we are able to implement and solve pieces of the analytics problem on several components simultaneously.

• For example, our People Counter feature utilized video analytics on the our camera to identify human traffic coming through the door and generate raw ‘metadata’ which it then records to the MultiSight Gateway recorder.  This analytics metadata on the recorder is further processed to generate people counts on a periodic basis.  The data is pushed to our cloud based services where it is stored, and further aggregated and summarized when reports are generated.
• When we released Heatmaps, we similarly deployed code to our cameras that summarized motion information in a specific way and transmitted that data to our recorder.  When our clients request a heatmap through our cloud-based API, the recorder processes that historical information and generates a tailored heatmap overlayed onto a visual reference scene.

Please log in or register below to read the full article.