Job title: rater – crop classification in satellite and street view imagesproject goal
the objective of this project is to classify crop types from satellite imagery by leveraging high-quality crop labels derived from street-view images of fields, providing a scalable ground-truth dataset for agricultural research and ai training.
Responsibilities
review satellite and street-view images to determine crop type or classify areas as uncultivated based on visual cues.
Follow established workflows and annotation protocols to ensure consistent and accurate labeling.
Apply domain knowledge to identify crop types even under partial occlusion.
Maintain accuracy, attention to detail, and consistency across large volumes of images.
Workflow
agricultural area presence – confirm if agricultural fields occupy more than a defined percentage (e.g., 40%) of the image view.
Field visibility assessment – evaluate visibility of the field:
partially occluded but identifiable – annotate the crop type or mark as uncultivated.
Clearly visible and identifiable – proceed to assign the correct crop label.
Qualifications
academic background: bachelor’s degree (or higher) in agriculture, agronomy, crop science, agricultural engineering, horticulture, or a related field.
Hands-on experience: knowledge in geography, remote sensing, environmental science, or gis with exposure to crop identification.
Agriculture experience: previous involvement in crop identification, agricultural surveys, or image annotation.
Visual identification skills: ability to distinguish crop types from both partial and full views.
Preferred skills
strong attention to detail and familiarity with diverse crop types.
Prior experience using image annotation tools and platforms is an advantage.
Ability to work independently and deliver results under defined timelines.