Rosaly Valwater Lettuce Detection

Project overview

Agricultural images from the Rosaly Valwater field environment were annotated for a computer vision dataset built around the detection of lettuce plants. Individual plants were identified and enclosed in accurate bounding boxes to produce high-quality training data for an AI-based object detection model.

The main objective was to prepare a reliable dataset for automatically detecting lettuce plants under real-world agricultural conditions.

Work performed

• Reviewed and organised the agricultural image dataset before annotation.
• Identified individual lettuce plants throughout the field images.
• Drew accurate bounding boxes around each visible lettuce plant.
• Applied the Lettuce class consistently across the whole dataset.
• Annotated partially visible lettuce plants where applicable.
• Distinguished lettuce carefully from surrounding weeds, soil and other vegetation.
• Handled lettuce plants at a range of growth stages and sizes.
• Performed quality checks and corrected inaccurate, missing or poorly positioned bounding boxes.
• Ensured the annotations were consistent and suitable for object detection model training.

What made it hard

• Differentiating lettuce plants from surrounding weeds and vegetation.
• Small lettuce plants at early growth stages were difficult to detect.
• Closely positioned and partially overlapping plants required careful placement.
• Plant sizes and growth stages varied widely across the dataset.
• Complex agricultural backgrounds contained soil, weeds and other crops.
• Varying lighting conditions and shadows affected plant visibility.
• Keeping bounding boxes tight and accurate around individual plants demanded sustained attention.

Result

A high-quality Rosaly Valwater lettuce detection dataset was completed with accurate bounding box annotations. The dataset was prepared for training a computer vision model capable of automatically detecting lettuce plants in real-world agricultural field environments.

Final deliverable

Accurate and consistent bounding box annotations for the Lettuce class were delivered across the dataset. The final output was prepared to the project requirements and was ready for AI model training and automated agricultural crop detection applications.