POA Segmentation

Project overview

The POA Segmentation project involved annotating agricultural field images to identify and precisely segment POA (Poa) plants. The dataset was prepared for computer vision model training focused on accurate plant segmentation and agricultural weed monitoring.

Work performed

• Reviewed agricultural field images and identified POA plants.
• Precise polygon segmentation masks were created around individual plants.
• Each mask followed the visible boundaries of the plant closely.
• POA was differentiated from surrounding crops, weeds, soil and vegetation.
• Overlapping and partially visible plants were handled individually.
• Consistent segmentation quality was maintained throughout the dataset.
• Inaccurate or incomplete masks were reviewed and corrected.
• Prepared the dataset for plant segmentation model training.

What made it hard

• Distinguishing POA from visually similar grasses and surrounding vegetation.
• Dense and complex field backgrounds made precise boundaries hard to place.
• Overlapping plants and partially visible leaves complicated mask drawing.
• Small or irregularly shaped plants were difficult to segment accurately.
• Polygon boundaries had to stay consistent across varied image conditions.

Result

A high-quality segmented dataset was prepared for POA plant segmentation, suitable for training computer vision models for agricultural weed identification, segmentation and monitoring.

Final deliverable

The final dataset contained accurate and consistent polygon segmentation masks for the POA class and was prepared for computer vision model training and agricultural weed monitoring applications.