Starting the Flask Backend:

Run the following in yolov5-deepsort:python web-ui/app.py. The frontend react app is in https://github.com/ykcai-daniel/ui-lost-and-found. Alternatively, you can also start a local opencv demo (make sure the paths to images and videos are correct): python web-ui/backend.py.

When the target person is located, its bounding box turn green: At the moment when the person losts his or her belonging, "Lost" will pop up on the bounding box of the suitcase (Currently only suitcase is supported). The suitcase is "lost" when pixel distance between the center of the person and the suitcase exceed a threshold. However, this might not be the perfect criteria and we might need some better distance estimation algorithms.

video capture

\demo.py\main

frame of lost

\tracker.py\plot_bboxes

where to add face id

\tracker.py\plot_bboxes\image

where to change suitcase to other items

\tracker.py\demo.py -> item_to_detect