Shoplifters are one of the biggest troublemaker for supermarket and malls. Monitoring shoplifters continuously is a quite tough task. This poc is related to automation in shoplifting detection using Deep Learning .This poc will do the work for you, by keeping eye on store via live CCTV footage.
- We trained our model with various supermarket’s CCTV footage
- Our system works on live CCTV stream
- It works in video frames which makes it efficient for model to trace and analyse activity
- With help of Transfer Learning, we are able to identify human’s visual-skeleton
- Once human is identified it predicts it pose using Computer Vision, based on visual-skeleton
- A rectangular box is visible on screen which bind human on it.
- Later on, it predicts it behaviour by analysing human’s behaviour.
- This works as segment and shows human behaviour above that rectangular box.
- It does same for every person present there.
- Once we identify its behaviour we can keep eye on such entities and prevent shoplifting
Later on, it predicts it behavior by analyzing human’s behavior this works as segment and shows human behavior above that rectangular box It does same for every person present there Once we identify its behavior we can keep eye on such entities and prevent shoplifting.
Key values / Achievements
- Transfer learning helped the model to detect the pose of human in form of visual skeleton.
- The model was highly capable enough to detect the thief from the video capture.
- The image shown is the screenshot taken from the video captured.