MorseDecoder

This project uses Python and OpenCv to decode Morse Code messages transmitted through flashes(bursts) of light.

Learn more about morse code from here : https://en.wikipedia.org/wiki/Morse_code Learn more about morse code tree representation from here : http://www.cranburyscouts.org/MorseTree.htm

The idea behind this project is to help navy or ships in general to detect flashes of light originating from any location at night time, and decode the morse code underneath it.

CURRENT LIMITATION - Since this project was originally intended to operate at night time, the code works on a relatively darker background with one bright source of illumination.

OpenCV is used detect the location of light source present in a scene. After the location has been determined, intermediate time units for both light_time (i.e. the time or the number of frames for which the light source was ON) and dark_time (i.e. the time or the number of frames for which the light source was OFF) are used to detect dots,dashes, inter-word pause and inter-sentence pause.

The morse code sequence is then decoded using the classic morse code tree represenattion.

Package requirements

1. NumPy
2. OpenCV

HOW TO USE:

There two functions in main.py file depending upon your use case:

1. Inside readFromFile():

Mention the path of the video file to path variable.

---Provide threshold values as input----

Provide value to light_min_thresh variable (this is 1 unit of light ON time. And this value will vary depending upon your video.)

Provide values to dark_min_thresh, dark_med_thresh, dark_max_thresh (these are units of light OFF time. These stand for inter-charcater gap. inter-word gap and inter-sentence gap respectively.And these values will vary depending upon your video. )

If you are not sure what the threshold values should be, keep the default values and execute the code. And the observe the output of "light times" list and "dark times" list. It will give you an idea for the threshold values. Provide the updated threshold values and re execute teh code to get the correct morse code message and the decoded message.

2. Inside readFromLiveFeed():

Mention the path of the video file to path variable.

This function plays the video in a loop and performs sampling on "light ON" and "light OFF" times by analyzing each frame of teh video and maintaining a counter.

The correct decoded message will be displayed in output, once the function has found the optimal threshold values for light and dark times.

FUTURE WORK

1. Real time camera feed Instead of video frames, real time elapsed in microseconds will be considered.

2. This project has been developed under the assumption that light houses and other naval vehicles use a light source with fixed light ON and light OFF time durations. (or the light ON and light OFF time durations reach eventual consistency.) But this may not work if the time durations vary too often. (Ex: A lone survivor stranded on an island and using a falshlight) Future work may include a ML algorithm which will handle the above mentioned case.