8 Big Issues with Traditional People Counting

What’s broken? 

‍

1. SINGLE CAMERA PROBLEM – Environment optimized for a single (type of) camera, making counting difficult

2. JANITOR PROBLEM – A person, such as a janitor, walking through an area gets counted multiple times over different cameras

3. OVERLAP PROBLEM – Related to the janitor problem, if a set of cameras point at the same space, viewpoints overlap and people can get counted multiple times (unless they’re de-duped) 

4. TECHNICAL ARCANA – Technical setup (camera angles, positions) that business analysts must deal with to get the metrics they’re actually looking for

5. ON-PREM HARDWARE – Traditionally, expensive hardware was required to get accurate counting metrics

6. INACCURATE CLOUD ANALYSIS – Unsophisticated cloud solutions can be inaccurate as well, since a lot of data must be transported tot he cloud to get the best results

7. INSIGHTS – Many solutions don’t provide the fine-grained tools to parse and compare data

8. NO HOLISTIC VIEW – Analysts’ questions are different than a security guard’s questions; typically there’s been no holistic view you can solve for