Sorghum Panicle Detection

Designing Computer Vision Models for Detecting Sorghum Panicles in Large Field Scans

Sorghum Field

Collaborators: Pauli Lab, University of Queensland

  • Architectured CNNs for crop-level panicle detection on Field Scanner Data
  • Increased testing speeds by 30% via developing a randomized sample Python script
  • Led the team effort of manual test data labeling for over 400 sample scans in a week