Introduction

This is a research project under MORSE Studio Georgia State University under Dr. Ashwin Ashok
The project aims to understand feasibility of using mmWave Radar setup to detect tiny drone imitating a fly/mosquito with the help of range, doppler and sound signatures of the object of interest
Special Thanks to Argha Sen, IIT Kharagpur, for allowing us to refer https://github.com/arghasen10/mmHER for the raw data streaming and parsing code using DCA1000EVM

---

Hardware Setup

Step 1: Make sure the IWR6843ISK (mmWave board) is connected with the mmWAVEICBOOST and DCM1000EVM along with the swicthes as follows

Step 2: Make sure you have flashed the DCA1000EVM with the apt FPGA image using the Lattice Diamond Programmer.
See document: https://www.ti.com/lit/pdf/spruij4 under FPGA - SPI Flash Programming Mode

---

mmWave Studio Configuration Setup

Step 1: Run the dataCaptureScript.lua from the repository in the mmWave Studio

Step 2: Once the script is done running, we are ready to capture data using the datacapture.py script.

---

Recording Data

Run the script with the desired duration of data capture in minutes as a command-line argument. This will create a .bin file in the same directory which can be later used for plotting and analysis. For example:

python data_capture.py 5

---

Plotting the Data

Using the created .bin file, to view the data frame by frame, with going to the next frame using the key 'q', use the following usage
We set the 3rd argument as 0 indicating we do not need to capture a video of the results.

python plotData.py <filename.bin> 0 

The results should be as follows:<br>

Please adjust the vmin and vmax values of imshow() as per your data to focus on specific signatures you need in the plotting part of the plot_doppler_range_power() function.
Also you can set the number of frames to be viewed by changing the NumOfFrames variable in the main()

---

Running the Plot Results as a video

For showing and presenting the results its inefficient to go frame by frame especially when you can have thousands of frames. It's better to create a video of the plots especially to observe the doppler.
By setting the 3rd argument as 1 , we save images of the plots frame by frame, by creating a output_images directory.
These images are then stiched as a video using the OpenCv VideroWriter and available as an '.mp4' file in the same directory as 'Plot_Results.mp4'.

python plotData.py <filename.bin> 1 

---

Link to the Experiments Data

https://studentgsu-my.sharepoint.com/:f:/r/personal/sdeshpande8_student_gsu_edu/Documents/MmWave_Experiments?csf=1&web=1&e=hezIL6