How happy95059/computer_vision_hw1 works
Project type and purpose
• This is a small, modular, Python-based computer-vision toolset (a “homework1” assignment) that processes short video clips to detect and replace objects (coin, poker card, dice) against a predicted background.
File structure
• Top-level:
• 61175069H_a1/source/
– hw1-1/
· get_background.py (background-capture utility)
· hw1-1_coint.py (main coin-removal script)
– hw1-2/
· poker_matting.py (card matting helper)
· hw1-2_poker.py (main card-replacement script)
– hw1-3/
· dice_matting.py (dice matting helper)
· hw1-3_dice.py (main dice-replacement script)
– test.py (generic test harness for all HWI scripts)
• Three .mp4 videos (hw1-1.mp4, hw1-2.mp4, hw1-3.mp4) used as inputs
• No external microservices or databases; all is on local file I/O and in-memory OpenCV processing.
README for additional insights
• Lists custom functions (get_hand_direction, get_origin_color, change_color2background, get_coint_direction, show_edge) that form the reusable CV library.