Skip to content

FedericoPivotto/billiard-video-analysis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Sport video analysis for billiard matches

Final project - June + July 2024.

Authors

  • Federico Pivotto, 2121720
  • Fabrizio Genilotti, 2119281
  • Leonardo Egidati, 2106370

Instructions

Compilation

To compile the computer vision system you have to:

  1. Create the /build directory in the root directory of the repository;
  2. Enter in the /build directory;
  3. Run the command cmake .. from terminal;
  4. Build the project running the command make.

Execution

To run the built computer vision system you have to:

  1. Enter in the /build directory;
  2. Run the command ./system from terminal.

Output

At the end of the execution of the computer vision system, the generated outputs are saved in the /result directory.

For each video sub-directory /gameX_clipY included in /dataset is generated an output directory /gameX_clipY in the /result directory, containing the following sub-directories:

  • /frames: first and last frame of the video;
  • /border_detection: first and last frame with border detection of the playing field;
  • /object_detection: first and last frame with ball detection and classification;
  • /bounding_boxes: text file with list of ball bounding boxes detected in the first and last frame;
  • /segmentation: first and last frame with segmentation of the playing field;
  • /masks: first and last frame with mask segmentation of the playing field and surrounding environment;
  • /output: first and last frame given by the union of border detection and segmentation;
  • /minimap: first and last frame with 2D top-view visualization map;
  • /metrics: mAP and mIoU metrics measured on the first and last frame;
  • Video with 2D-top-view visualization map.