OpenCV-DAM-System

Introduction

This project emulates a Drosophila Activity Monitoring (DAM) system using OpenCV and a low spec 1 FPS camera. It was commissioned by a sleep study lab in University of Tsukuba, Japan. Its is able to analyze both a camera stream and a video file passed to it as arguments.

Read about what a DAM system is here: https://blogs.brandeis.edu/flyonthewall/fly-life-watching-fruit-flies-sleep/

Video used to test program: https://drive.google.com/file/d/1qktrlR4CitGu-efn7M3S3IUm1zbLW9oH/view?usp=sharing

How to use:

1. Call using command line arguments:

• Flycircadian.py ( PATHTOFILE | cameraID ) ( int FramesToRun) ( CSVFilename )

• PATHTOFILE: should be the path to file or just the filename if in same directory

• cameraID: should be an integer ID of the camera connected to the laptop if reading from camera

• FramesToRun: Frames to run program, should be "endless" if running indefinitely

• CSVFilename: Name of csv filename to save to, program will append '.csv' to end automatically

EXAMPLE FOR VIDEO STREAM: Flycircadian.py Tests/flyvideo1.avi 3600 flys.csv

EXAMPLE FOR CAMERA STREAM: Flycircadian.py 0 endless flys.csv

WARNING: Program will overwrite csv files with filename in directory without warning, please move csv files out of directory after each experiment

2. Partition flies in initial frame by dragging boxes with mouse and labeling in popup UI

• Strange results may occur if any of the boxes overlap
• Please do not click cancel on labeling window for now

3. Let camera run with program for the duration of the experiment, avoid disturbing lighting condition

• Frames to set flies to asleep can be changed by changing value of TIMEOUT inside Flycircardian.py

4. Program will output results into .csv format

• Program will only save to a csv file every 5 new entrys to the log by default