Getting Started with Computer Vision - Carleton-SRCL/SPOT GitHub Wiki
Welcome to SPOT Computer Vision!
Computer vision is a branch of artificial intelligence involving digital systems that can detect and process visual information. It is crucial for the detection and tracking of specific objects in digital images, videos, and other elements because it teaches computers how to identify and analyze images so that they can interpret them and perform actions or propose recommendations based on what they see. In this case, the platforms on the Spacecraft Proximity Operations Testbed detect "target" and "obstacle" platforms using the ZED Camera.
SETUP Procedure
This guide will help install all required coding languages, IDEs, and libraries necessary for Computer Vision.
Developed Functions
These are the developed computer vision functions.
Single Camera
- Single Camera Calibration (CalibrationProcedure.py)
- Single Camera, Single ArUco Marker Pose Detection (Live_Pose_Det.py)
Stereo Camera (Single ArUco Marker)
- ZED Camera Calibration (stereovision_calibration.py)
- ZED Camera, Single ArUco Marker Pose Detection (stereovision.py)
Stereo Camera (Multiple ArUco Markers)
- Optimized ZED Camera Calibration (CCS.py)
- ZED Camera, Multiple ArUco Marker Pose Detection (BFD_RPY_Stereo.py)
- ZED Camera, Multiple ArUco Marker Pose Detection with Varying ArUco Size (MultisizeAruco_CVIS_PPL_Integration_v6.py)
- ZED Camera Finalized ArUco Pose Estimation (MULTI_AR_VIS_BASE_WO.py)
Object Detection
- Adaptive Object Detection and Image Processing (ADPT_HA.py)
- Obstacle Detection and Image Processing (ODC_SS.py)
- Obstacle Distance Determination through Depth Mapping (SDM_AC.py)
Final Code
Functions MULTI_AR_VIS_BASE_WO.PY, ODC_SS.py, SDM_AC.py, _ADPT_HA.py_ cooperatively generate the final Computer Vision output sent to the Path Planning subteam.