Project Introduction

Pet CCTV is back large and in charge.


Pet cameras are becoming increasingly popular as the technology becomes more accessible and people find themselves spending more time away from their home and pets. At their core, traditional approaches to pet monitoring are all quite similar – a camera which watches 24/7, and transmits this feed to a pet owner’s personal device. However, this means that the pet owner would need to be actively watching (or spend time reviewing footage later) in order to get any idea of what their pet has been up to. In this project, we hope to add more convenience to the concept of a pet monitoring system by incorporating computer vision and machine learning to detect bad behaviors as they happen, and to provide a concise report of the animal’s activity. It is our belief that this system could prove valuable in any home with pets. We will create a system to monitor one room of the house – ideally with the potential that multiple systems could be linked together for more complete coverage of a multi-room house – that can track the location of an animal when it is in frame, differentiate between multiple animals, determine if an animal goes into a user-defined forbidden area, and report all of this data to the user in a neat and comprehensive web portal.