Dataset for "Computer Vision for Resource Optimisation in Dairy Cows Drinking and Backyard Chickens Nesting behaviour"

Animal Behavior Analysis with Computer Vision

This repository contains two YOLO-based computer vision projects for animal behavior analysis, along with training datasets and processed results:

Datasets

## Poultry Behaviour and Egg Detection DatasetContains 1272 high-resolution (2560x1440) images with 1172 annotation files in YOLO format for training models to detect:- Chickens- Guinea Fowls- Eggs

Images were captured during summer months from a backyard coop, specifically for analyzing nest occupancy patterns and egg-laying behavior.

Directory Structure:

training_data_(chicken_guinneafowl_egg)/

train/

images/

labels/

val/

images/

labels/

labels.txt

## Cattle Feeding Behaviour DatasetContains 1140 images (1920x1080) with matching annotation files for training YOLO models to detect:- Cow Heads- Cow Bodies- Water Troughs

This dataset was derived from research conducted by Koskela et al. (2023): https://doi.org/10.23986/afsci.111665

Directory Structure:

training_data_(head_body_trough)/

train/

images/

labels/

val/

images/

labels/

labels.txt

Complementary_files include:

## Poultry Analysis- 22 CSV files containing temporal bird occupancy data and egg count statistics across 11 days of monitoring- Detailed nest usage patterns and bird preferences

## Cattle Analysis- Processed video tracking data showing cow feeding interactions- Individual animal tracking information with temporal and spatial dimensions

## Python scripts (coop_analysis.py and cow_interaction_analysis.py), associated README files, and requirements.txt specifications

These materials were developed as part of research demonstrating practical applications for automated monitoring of animal behavior in poultry and cattle environments, conducted for the thesis "Computer Vision for Resource Optimisation in Dairy Cows Drinking and Backyard Chickens Nesting behaviour" (DOI: ).