Computer Vision Developer Required – Security Monitoring

We operate a remote video monitoring centre in New Zealand, reviewing CCTV alarms generated by motion detection systems. When an alarm occurs, our operators open a web-based video player in Google Chrome and review footage from a few seconds before the alarm through to approximately 30 seconds afterwards. The challenge is that many activations are caused by very small or difficult-to-see movement, including: Mice and rodents Birds Insects Plastic bags or rubbish Vegetation movement Lighting changes Small distant human movement These can be difficult for operators to identify in real time, particularly at night or in poor lighting conditions. We are looking for an experienced Computer Vision Developer to build a Windows application that helps operators visually identify motion events faster and more accurately. Desired Solution We envisage a lightweight Windows desktop application that runs over the top of, or alongside, Google Chrome while video footage is being viewed. The software should: Monitor a defined area of the screen displaying CCTV footage Detect motion occurring within the video Visually highlight detected movement Draw bounding boxes around moving objects Make subtle motion more obvious to operators Operate in near real-time while footage is being replayed The goal is not to automate incident decisions. The goal is to improve human detection of subtle movement that may otherwise be missed. Initially we are looking for basic cost estimates to gather further requirements, and build a working proof of concept. We also need a cost estimate to take the proof of concept into a finished application that we can rely on hundreds of times a day. In the future we may explore teaching the AI not to direct our attention to certain things, i.e. vegetation.