Track the markers on the Zozo-suit with OpenCV.
python3 -m venv env
source env/bin/activate
pip install -r requirements.txt
Run detect_points.py
with a filename as first argument.
Example:
python3 detect_points.py IMG_20180911_163153_c.jpg
Runnig the script will create two files:
point_positions.png
The original image with positions, IDs and distances drawn into the image.collected_points.png
A usually very wide image with all points in their original skewed position, the corrected version and an image with a drawing of the point it detected. Below is the ID, the confidence score in percent and the calculated distance.
If you want to use detect_points.py
in your own code,
import detect_points
. The function does all the heavy
lifting and returns IDs, confidence scores (0 to 1),
positions and distances. If you need raw data from the
recognition use the raw_data
dictionary which is returned
as the last variable.