[TASK] Object Detection Benchmarking

[jorgenfj]

Description of task

Robust object detection using deep learning, like a yolo model, is time consuming to implement. Therefore we need an easy to implement alternative for object detection to use as fallback.

Recommended alternatives:

• SIFT
• ORB
• SURF

Things to consider

• Speed
• Accuracy
• Resources (CPU or GPU)
• Implementation Difficulty

Suggested Workflow

• Read up on relevant algorithm
  Comparison of SIFT, SURF and ORB
  OpenCV comparison
  Object Tracking with SIFT Algorithm using OpenCV
  Detecting and Tracking Objects with ORB Algorithm using OpenCV
  SIFT Feature Detection Developed in ROS1 by vortex
• Implement in ROS2

Specifications

• Should be able to detect valve for TACC mission
• Moderate accuracy is acceptable
• Should output a bounding box for detected object

Contacts

@jorgenfj

Code Quality

• Every function in header files are documented (inputs/returns/exceptions)
• The project has automated tests that cover MOST of the functions and branches in functions (pytest/gtest)
• The code is documented on the wiki (provide link)